Cohere Model Release (#29622)
* Cohere Model Release (#1) Cohere Model Release * Remove unnecessary files and code (#2) Some cleanup * Delete cohere-model directory (#3) * Make Fix (#5) * Pr fixes (#6) * fixes for pr * pr fixes for the format * pr fixes for the format * src/transformers/models/auto/tokenization_auto.py * Tokenizer test (#8) * tokenizer test * format fix * Adding Docs and other minor changes (#7) * Add modeling tests (#9) * Smol Fix (#11) * tokenization tests are fixed * format fixes * fix pr doc tests * fix pr doc tests * fix pr doc tests * fix pr style check * small changes in cohere.md * FIX: Address final comments for transformers integration (#13) * fix modeling final nits and add proper test file * for now leave empty tests * add integration test * push new test * fix modeling cohere (#14) * Update chat templates to use the new API (#15) --------- Co-authored-by: ahmetustun <ahmetustun89@gmail.com> Co-authored-by: Younes Belkada <49240599+younesbelkada@users.noreply.github.com> Co-authored-by: Matt <Rocketknight1@users.noreply.github.com>
This commit is contained in:
parent
53d891247b
commit
0e4a1c3401
|
@ -331,6 +331,7 @@ Current number of checkpoints: ![](https://img.shields.io/endpoint?url=https://h
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -327,6 +327,7 @@ Aktuelle Anzahl der Checkpoints: ![](https://img.shields.io/endpoint?url=https:/
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -304,6 +304,7 @@ Número actual de puntos de control: ![](https://img.shields.io/endpoint?url=htt
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -325,6 +325,7 @@ Nombre actuel de points de contrôle : ![](https://img.shields.io/endpoint?url=h
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** publié dans l'article [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) par James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (de Salesforce) publié dans l'article [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) par Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (de MetaAI) publié dans l'article [Code Llama : Modèles ouverts fondamentaux pour le code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) par Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (de Cohere) publié dans l'article [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) parCohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (de Microsoft Research Asia) publié dans l'article [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) par Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (de YituTech) publié dans l'article [ConvBERT : Amélioration de BERT avec une convolution dynamique basée sur des plages](https://arxiv.org/abs/2008.02496) par Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (de Facebook AI) publié dans l'article [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) par Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -278,6 +278,7 @@ conda install conda-forge::transformers
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (सेल्सफोर्स से) साथ में पेपर [प्रोग्राम सिंथेसिस के लिए एक संवादात्मक प्रतिमान](https://arxiv.org/abs/2203.13474) एरिक निजकैंप, बो पैंग, हिरोआकी हयाशी, लिफू तू, हुआन वांग, यिंगबो झोउ, सिल्वियो सावरेस, कैमिंग जिओंग रिलीज।
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (MetaAI से) Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve. द्वाराअनुसंधान पत्र [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) के साथ जारी किया गया
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (Cohere से) Cohere. द्वाराअनुसंधान पत्र [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) के साथ जारी किया गया
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (माइक्रोसॉफ्ट रिसर्च एशिया से) कागज के साथ [फास्ट ट्रेनिंग कन्वर्जेंस के लिए सशर्त डीईटीआर](https://arxiv.org/abs/2108.06152) डेपू मेंग, ज़ियाओकांग चेन, ज़ेजिया फैन, गैंग ज़ेंग, होउकियांग ली, युहुई युआन, लेई सन, जिंगडोंग वांग द्वारा।
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (YituTech से) साथ में कागज [ConvBERT: स्पैन-आधारित डायनेमिक कनवल्शन के साथ BERT में सुधार](https://arxiv.org/abs/2008.02496) जिहांग जियांग, वीहाओ यू, डाकान झोउ, युनपेंग चेन, जियाशी फेंग, शुइचेंग यान द्वारा।
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (Facebook AI से) साथ वाला पेपर [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) ज़ुआंग लियू, हेंज़ी माओ, चाओ-युआन वू, क्रिस्टोफ़ फीचटेनहोफ़र, ट्रेवर डेरेल, सैनिंग ज़ी द्वारा।
|
||||
|
|
|
@ -338,6 +338,7 @@ Flax、PyTorch、TensorFlowをcondaでインストールする方法は、それ
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (Salesforce から) Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong から公開された研究論文: [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474)
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (MetaAI から) Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve. から公開された研究論文 [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/)
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (Cohere から) Cohere. から公開された研究論文 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>)
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (Microsoft Research Asia から) Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang から公開された研究論文: [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152)
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (YituTech から) Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan から公開された研究論文: [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496)
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (Facebook AI から) Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie から公開された研究論文: [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545)
|
||||
|
|
|
@ -253,6 +253,7 @@ Flax, PyTorch, TensorFlow 설치 페이지에서 이들을 conda로 설치하는
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (Salesforce 에서) Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong 의 [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) 논문과 함께 발표했습니다.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (MetaAI 에서 제공)은 Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.의 [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/)논문과 함께 발표했습니다.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (Cohere 에서 제공)은 Cohere. 의 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>)논문과 함께 발표했습니다.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (Microsoft Research Asia 에서) Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang 의 [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) 논문과 함께 발표했습니다.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (YituTech 에서) Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan 의 [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) 논문과 함께 발표했습니다.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (Facebook AI 에서) Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie 의 [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) 논문과 함께 발표했습니다.
|
||||
|
|
|
@ -336,6 +336,7 @@ Número atual de pontos de verificação: ![](https://img.shields.io/endpoint?ur
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -326,6 +326,7 @@ conda install conda-forge::transformers
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -328,6 +328,7 @@ Flax, PyTorch లేదా TensorFlow యొక్క ఇన్స్టా
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -327,6 +327,7 @@ Số lượng điểm kiểm tra hiện tại: ![](https://img.shields.io/endpoi
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** được phát hành với bài báo [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (từ Salesforce) được phát hành với bài báo [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (từ MetaAI) được phát hành với bài báo [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (từ Cohere) được phát hành với bài báo [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (từ Microsoft Research Asia) được phát hành với bài báo [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (từ YituTech) được phát hành với bài báo [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (từ Facebook AI) được phát hành với bài báo [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -277,6 +277,7 @@ conda install conda-forge::transformers
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (来自 Salesforce) 伴随论文 [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) 由 Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong 发布。
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (来自 MetaAI) 伴随论文 [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) 由 Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve 发布。
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (来自 Cohere) 伴随论文 [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) 由 Cohere 发布。
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (来自 Microsoft Research Asia) 伴随论文 [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) 由 Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang 发布。
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (来自 YituTech) 伴随论文 [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) 由 Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan 发布。
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (来自 Facebook AI) 伴随论文 [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) 由 Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie 发布。
|
||||
|
|
|
@ -289,6 +289,7 @@ conda install conda-forge::transformers
|
|||
1. **[CLVP](https://huggingface.co/docs/transformers/model_doc/clvp)** released with the paper [Better speech synthesis through scaling](https://arxiv.org/abs/2305.07243) by James Betker.
|
||||
1. **[CodeGen](https://huggingface.co/docs/transformers/model_doc/codegen)** (from Salesforce) released with the paper [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong.
|
||||
1. **[CodeLlama](https://huggingface.co/docs/transformers/model_doc/llama_code)** (from MetaAI) released with the paper [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) by Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, Gabriel Synnaeve.
|
||||
1. **[Cohere](https://huggingface.co/docs/transformers/main/model_doc/cohere)** (from Cohere) released with the paper [Command-R: Retrieval Augmented Generation at Production Scale](<https://txt.cohere.com/command-r/>) by Cohere.
|
||||
1. **[Conditional DETR](https://huggingface.co/docs/transformers/model_doc/conditional_detr)** (from Microsoft Research Asia) released with the paper [Conditional DETR for Fast Training Convergence](https://arxiv.org/abs/2108.06152) by Depu Meng, Xiaokang Chen, Zejia Fan, Gang Zeng, Houqiang Li, Yuhui Yuan, Lei Sun, Jingdong Wang.
|
||||
1. **[ConvBERT](https://huggingface.co/docs/transformers/model_doc/convbert)** (from YituTech) released with the paper [ConvBERT: Improving BERT with Span-based Dynamic Convolution](https://arxiv.org/abs/2008.02496) by Zihang Jiang, Weihao Yu, Daquan Zhou, Yunpeng Chen, Jiashi Feng, Shuicheng Yan.
|
||||
1. **[ConvNeXT](https://huggingface.co/docs/transformers/model_doc/convnext)** (from Facebook AI) released with the paper [A ConvNet for the 2020s](https://arxiv.org/abs/2201.03545) by Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, Saining Xie.
|
||||
|
|
|
@ -310,6 +310,8 @@
|
|||
title: CodeGen
|
||||
- local: model_doc/code_llama
|
||||
title: CodeLlama
|
||||
- local: model_doc/cohere
|
||||
title: Cohere
|
||||
- local: model_doc/convbert
|
||||
title: ConvBERT
|
||||
- local: model_doc/cpm
|
||||
|
|
|
@ -95,6 +95,7 @@ Flax), PyTorch, and/or TensorFlow.
|
|||
| [CLVP](model_doc/clvp) | ✅ | ❌ | ❌ |
|
||||
| [CodeGen](model_doc/codegen) | ✅ | ❌ | ❌ |
|
||||
| [CodeLlama](model_doc/code_llama) | ✅ | ❌ | ✅ |
|
||||
| [Cohere](model_doc/cohere) | ✅ | ❌ | ❌ |
|
||||
| [Conditional DETR](model_doc/conditional_detr) | ✅ | ❌ | ❌ |
|
||||
| [ConvBERT](model_doc/convbert) | ✅ | ✅ | ❌ |
|
||||
| [ConvNeXT](model_doc/convnext) | ✅ | ✅ | ❌ |
|
||||
|
|
|
@ -0,0 +1,141 @@
|
|||
# Cohere
|
||||
|
||||
## Overview
|
||||
|
||||
The Cohere Command-R model was proposed in the blogpost [Command-R: Retrieval Augmented Generation at Production Scale](https://txt.cohere.com/command-r/) by the Cohere Team.
|
||||
|
||||
The abstract from the paper is the following:
|
||||
|
||||
*Command-R is a scalable generative model targeting RAG and Tool Use to enable production-scale AI for enterprise. Today, we are introducing Command-R, a new LLM aimed at large-scale production workloads. Command-R targets the emerging “scalable” category of models that balance high efficiency with strong accuracy, enabling companies to move beyond proof of concept, and into production.*
|
||||
|
||||
*Command-R is a generative model optimized for long context tasks such as retrieval augmented generation (RAG) and using external APIs and tools. It is designed to work in concert with our industry-leading Embed and Rerank models to provide best-in-class integration for RAG applications and excel at enterprise use cases. As a model built for companies to implement at scale, Command-R boasts:
|
||||
- Strong accuracy on RAG and Tool Use
|
||||
- Low latency, and high throughput
|
||||
- Longer 128k context and lower pricing
|
||||
- Strong capabilities across 10 key languages
|
||||
- Model weights available on HuggingFace for research and evaluation
|
||||
|
||||
Checkout model checkpoints [here](https://huggingface.co/CohereForAI/c4ai-command-r-v01).
|
||||
This model was contributed by [Saurabh Dash](https://huggingface.co/saurabhdash) and [Ahmet Üstün](https://huggingface.co/ahmetustun). The code of the implementation in Hugging Face is based on GPT-NeoX [here](https://github.com/EleutherAI/gpt-neox).
|
||||
|
||||
## Usage tips
|
||||
|
||||
<Tip warning={true}>
|
||||
|
||||
The checkpoints uploaded on the Hub use `torch_dtype = 'float16'`, which will be
|
||||
used by the `AutoModel` API to cast the checkpoints from `torch.float32` to `torch.float16`.
|
||||
|
||||
The `dtype` of the online weights is mostly irrelevant unless you are using `torch_dtype="auto"` when initializing a model using `model = AutoModelForCausalLM.from_pretrained("path", torch_dtype = "auto")`. The reason is that the model will first be downloaded ( using the `dtype` of the checkpoints online), then it will be casted to the default `dtype` of `torch` (becomes `torch.float32`), and finally, if there is a `torch_dtype` provided in the config, it will be used.
|
||||
|
||||
Training the model in `float16` is not recommended and is known to produce `nan`; as such, the model should be trained in `bfloat16`.
|
||||
|
||||
</Tip>
|
||||
The model and tokenizer can be loaded via:
|
||||
|
||||
```python
|
||||
# pip install transformers
|
||||
from transformers import AutoTokenizer, AutoModelForCausalLM
|
||||
|
||||
model_id = "CohereForAI/c4ai-command-r-v01"
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
||||
model = AutoModelForCausalLM.from_pretrained(model_id)
|
||||
|
||||
# Format message with the command-r chat template
|
||||
messages = [{"role": "user", "content": "Hello, how are you?"}]
|
||||
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
|
||||
## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Hello, how are you?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
|
||||
|
||||
gen_tokens = model.generate(
|
||||
input_ids,
|
||||
max_new_tokens=100,
|
||||
do_sample=True,
|
||||
temperature=0.3,
|
||||
)
|
||||
|
||||
gen_text = tokenizer.decode(gen_tokens[0])
|
||||
print(gen_text)
|
||||
```
|
||||
|
||||
- When using Flash Attention 2 via `attn_implementation="flash_attention_2"`, don't pass `torch_dtype` to the `from_pretrained` class method and use Automatic Mixed-Precision training. When using `Trainer`, it is simply specifying either `fp16` or `bf16` to `True`. Otherwise, make sure you are using `torch.autocast`. This is required because the Flash Attention only support `fp16` and `bf16` data type.
|
||||
|
||||
|
||||
## Resources
|
||||
|
||||
A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with Command-R. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.
|
||||
|
||||
|
||||
<PipelineTag pipeline="text-generation"/>
|
||||
|
||||
Loading FP16 model
|
||||
```python
|
||||
# pip install transformers
|
||||
from transformers import AutoTokenizer, AutoModelForCausalLM
|
||||
|
||||
model_id = "CohereForAI/c4ai-command-r-v01"
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
||||
model = AutoModelForCausalLM.from_pretrained(model_id)
|
||||
|
||||
# Format message with the command-r chat template
|
||||
messages = [{"role": "user", "content": "Hello, how are you?"}]
|
||||
input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
|
||||
## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Hello, how are you?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
|
||||
|
||||
gen_tokens = model.generate(
|
||||
input_ids,
|
||||
max_new_tokens=100,
|
||||
do_sample=True,
|
||||
temperature=0.3,
|
||||
)
|
||||
|
||||
gen_text = tokenizer.decode(gen_tokens[0])
|
||||
print(gen_text)
|
||||
```
|
||||
|
||||
Loading bitsnbytes 4bit quantized model
|
||||
```python
|
||||
# pip install transformers bitsandbytes accelerate
|
||||
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
|
||||
|
||||
bnb_config = BitsAndBytesConfig(load_in_4bit=True)
|
||||
|
||||
model_id = "CohereForAI/c4ai-command-r-v01"
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
||||
model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=bnb_config)
|
||||
|
||||
gen_tokens = model.generate(
|
||||
input_ids,
|
||||
max_new_tokens=100,
|
||||
do_sample=True,
|
||||
temperature=0.3,
|
||||
)
|
||||
|
||||
gen_text = tokenizer.decode(gen_tokens[0])
|
||||
print(gen_text)
|
||||
```
|
||||
|
||||
|
||||
## CohereConfig
|
||||
|
||||
[[autodoc]] CohereConfig
|
||||
|
||||
## CohereTokenizerFast
|
||||
|
||||
[[autodoc]] CohereTokenizerFast
|
||||
- build_inputs_with_special_tokens
|
||||
- get_special_tokens_mask
|
||||
- create_token_type_ids_from_sequences
|
||||
- update_post_processor
|
||||
- save_vocabulary
|
||||
|
||||
## CohereModel
|
||||
|
||||
[[autodoc]] CohereModel
|
||||
- forward
|
||||
|
||||
|
||||
## CohereForCausalLM
|
||||
|
||||
[[autodoc]] CohereForCausalLM
|
||||
- forward
|
||||
|
||||
|
|
@ -39,6 +39,7 @@ FlashAttention-2 is experimental and may change considerably in future versions.
|
|||
FlashAttention-2 is currently supported for the following architectures:
|
||||
* [Bark](https://huggingface.co/docs/transformers/model_doc/bark#transformers.BarkModel)
|
||||
* [Bart](https://huggingface.co/docs/transformers/model_doc/bart#transformers.BartModel)
|
||||
* [Cohere](https://huggingface.co/docs/transformers/model_doc/cohere#transformers.CohereModel)
|
||||
* [DistilBert](https://huggingface.co/docs/transformers/model_doc/distilbert#transformers.DistilBertModel)
|
||||
* [Gemma](https://huggingface.co/docs/transformers/model_doc/gemma#transformers.GemmaModel)
|
||||
* [GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode#transformers.GPTBigCodeModel)
|
||||
|
@ -172,6 +173,7 @@ PyTorch's [`torch.nn.functional.scaled_dot_product_attention`](https://pytorch.o
|
|||
|
||||
For now, Transformers supports SDPA inference and training for the following architectures:
|
||||
* [Bart](https://huggingface.co/docs/transformers/model_doc/bart#transformers.BartModel)
|
||||
* [Cohere](https://huggingface.co/docs/transformers/model_doc/cohere#transformers.CohereModel)
|
||||
* [GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode#transformers.GPTBigCodeModel)
|
||||
* [Falcon](https://huggingface.co/docs/transformers/model_doc/falcon#transformers.FalconModel)
|
||||
* [Gemma](https://huggingface.co/docs/transformers/model_doc/gemma#transformers.GemmaModel)
|
||||
|
|
|
@ -37,7 +37,7 @@ You can finetune other architectures for causal language modeling following the
|
|||
Choose one of the following architectures:
|
||||
|
||||
<!--This tip is automatically generated by `make fix-copies`, do not fill manually!-->
|
||||
[BART](../model_doc/bart), [BERT](../model_doc/bert), [Bert Generation](../model_doc/bert-generation), [BigBird](../model_doc/big_bird), [BigBird-Pegasus](../model_doc/bigbird_pegasus), [BioGpt](../model_doc/biogpt), [Blenderbot](../model_doc/blenderbot), [BlenderbotSmall](../model_doc/blenderbot-small), [BLOOM](../model_doc/bloom), [CamemBERT](../model_doc/camembert), [CodeLlama](../model_doc/code_llama), [CodeGen](../model_doc/codegen), [CPM-Ant](../model_doc/cpmant), [CTRL](../model_doc/ctrl), [Data2VecText](../model_doc/data2vec-text), [ELECTRA](../model_doc/electra), [ERNIE](../model_doc/ernie), [Falcon](../model_doc/falcon), [Fuyu](../model_doc/fuyu), [Gemma](../model_doc/gemma), [GIT](../model_doc/git), [GPT-Sw3](../model_doc/gpt-sw3), [OpenAI GPT-2](../model_doc/gpt2), [GPTBigCode](../model_doc/gpt_bigcode), [GPT Neo](../model_doc/gpt_neo), [GPT NeoX](../model_doc/gpt_neox), [GPT NeoX Japanese](../model_doc/gpt_neox_japanese), [GPT-J](../model_doc/gptj), [LLaMA](../model_doc/llama), [Mamba](../model_doc/mamba), [Marian](../model_doc/marian), [mBART](../model_doc/mbart), [MEGA](../model_doc/mega), [Megatron-BERT](../model_doc/megatron-bert), [Mistral](../model_doc/mistral), [Mixtral](../model_doc/mixtral), [MPT](../model_doc/mpt), [MusicGen](../model_doc/musicgen), [MVP](../model_doc/mvp), [OpenLlama](../model_doc/open-llama), [OpenAI GPT](../model_doc/openai-gpt), [OPT](../model_doc/opt), [Pegasus](../model_doc/pegasus), [Persimmon](../model_doc/persimmon), [Phi](../model_doc/phi), [PLBart](../model_doc/plbart), [ProphetNet](../model_doc/prophetnet), [QDQBert](../model_doc/qdqbert), [Qwen2](../model_doc/qwen2), [Reformer](../model_doc/reformer), [RemBERT](../model_doc/rembert), [RoBERTa](../model_doc/roberta), [RoBERTa-PreLayerNorm](../model_doc/roberta-prelayernorm), [RoCBert](../model_doc/roc_bert), [RoFormer](../model_doc/roformer), [RWKV](../model_doc/rwkv), [Speech2Text2](../model_doc/speech_to_text_2), [StableLm](../model_doc/stablelm), [Starcoder2](../model_doc/starcoder2), [Transformer-XL](../model_doc/transfo-xl), [TrOCR](../model_doc/trocr), [Whisper](../model_doc/whisper), [XGLM](../model_doc/xglm), [XLM](../model_doc/xlm), [XLM-ProphetNet](../model_doc/xlm-prophetnet), [XLM-RoBERTa](../model_doc/xlm-roberta), [XLM-RoBERTa-XL](../model_doc/xlm-roberta-xl), [XLNet](../model_doc/xlnet), [X-MOD](../model_doc/xmod)
|
||||
[BART](../model_doc/bart), [BERT](../model_doc/bert), [Bert Generation](../model_doc/bert-generation), [BigBird](../model_doc/big_bird), [BigBird-Pegasus](../model_doc/bigbird_pegasus), [BioGpt](../model_doc/biogpt), [Blenderbot](../model_doc/blenderbot), [BlenderbotSmall](../model_doc/blenderbot-small), [BLOOM](../model_doc/bloom), [CamemBERT](../model_doc/camembert), [CodeLlama](../model_doc/code_llama), [CodeGen](../model_doc/codegen), [Cohere](../model_doc/cohere), [CPM-Ant](../model_doc/cpmant), [CTRL](../model_doc/ctrl), [Data2VecText](../model_doc/data2vec-text), [ELECTRA](../model_doc/electra), [ERNIE](../model_doc/ernie), [Falcon](../model_doc/falcon), [Fuyu](../model_doc/fuyu), [Gemma](../model_doc/gemma), [GIT](../model_doc/git), [GPT-Sw3](../model_doc/gpt-sw3), [OpenAI GPT-2](../model_doc/gpt2), [GPTBigCode](../model_doc/gpt_bigcode), [GPT Neo](../model_doc/gpt_neo), [GPT NeoX](../model_doc/gpt_neox), [GPT NeoX Japanese](../model_doc/gpt_neox_japanese), [GPT-J](../model_doc/gptj), [LLaMA](../model_doc/llama), [Mamba](../model_doc/mamba), [Marian](../model_doc/marian), [mBART](../model_doc/mbart), [MEGA](../model_doc/mega), [Megatron-BERT](../model_doc/megatron-bert), [Mistral](../model_doc/mistral), [Mixtral](../model_doc/mixtral), [MPT](../model_doc/mpt), [MusicGen](../model_doc/musicgen), [MVP](../model_doc/mvp), [OpenLlama](../model_doc/open-llama), [OpenAI GPT](../model_doc/openai-gpt), [OPT](../model_doc/opt), [Pegasus](../model_doc/pegasus), [Persimmon](../model_doc/persimmon), [Phi](../model_doc/phi), [PLBart](../model_doc/plbart), [ProphetNet](../model_doc/prophetnet), [QDQBert](../model_doc/qdqbert), [Qwen2](../model_doc/qwen2), [Reformer](../model_doc/reformer), [RemBERT](../model_doc/rembert), [RoBERTa](../model_doc/roberta), [RoBERTa-PreLayerNorm](../model_doc/roberta-prelayernorm), [RoCBert](../model_doc/roc_bert), [RoFormer](../model_doc/roformer), [RWKV](../model_doc/rwkv), [Speech2Text2](../model_doc/speech_to_text_2), [StableLm](../model_doc/stablelm), [Starcoder2](../model_doc/starcoder2), [Transformer-XL](../model_doc/transfo-xl), [TrOCR](../model_doc/trocr), [Whisper](../model_doc/whisper), [XGLM](../model_doc/xglm), [XLM](../model_doc/xlm), [XLM-ProphetNet](../model_doc/xlm-prophetnet), [XLM-RoBERTa](../model_doc/xlm-roberta), [XLM-RoBERTa-XL](../model_doc/xlm-roberta-xl), [XLNet](../model_doc/xlnet), [X-MOD](../model_doc/xmod)
|
||||
|
||||
|
||||
|
||||
|
|
|
@ -292,6 +292,7 @@ _import_structure = {
|
|||
"CodeGenConfig",
|
||||
"CodeGenTokenizer",
|
||||
],
|
||||
"models.cohere": ["COHERE_PRETRAINED_CONFIG_ARCHIVE_MAP", "CohereConfig"],
|
||||
"models.conditional_detr": [
|
||||
"CONDITIONAL_DETR_PRETRAINED_CONFIG_ARCHIVE_MAP",
|
||||
"ConditionalDetrConfig",
|
||||
|
@ -1174,6 +1175,7 @@ else:
|
|||
_import_structure["models.clip"].append("CLIPTokenizerFast")
|
||||
_import_structure["models.code_llama"].append("CodeLlamaTokenizerFast")
|
||||
_import_structure["models.codegen"].append("CodeGenTokenizerFast")
|
||||
_import_structure["models.cohere"].append("CohereTokenizerFast")
|
||||
_import_structure["models.convbert"].append("ConvBertTokenizerFast")
|
||||
_import_structure["models.cpm"].append("CpmTokenizerFast")
|
||||
_import_structure["models.deberta"].append("DebertaTokenizerFast")
|
||||
|
@ -1817,6 +1819,7 @@ else:
|
|||
"CodeGenPreTrainedModel",
|
||||
]
|
||||
)
|
||||
_import_structure["models.cohere"].extend(["CohereForCausalLM", "CohereModel", "CoherePreTrainedModel"])
|
||||
_import_structure["models.conditional_detr"].extend(
|
||||
[
|
||||
"CONDITIONAL_DETR_PRETRAINED_MODEL_ARCHIVE_LIST",
|
||||
|
@ -5119,6 +5122,7 @@ if TYPE_CHECKING:
|
|||
CodeGenConfig,
|
||||
CodeGenTokenizer,
|
||||
)
|
||||
from .models.cohere import COHERE_PRETRAINED_CONFIG_ARCHIVE_MAP, CohereConfig
|
||||
from .models.conditional_detr import (
|
||||
CONDITIONAL_DETR_PRETRAINED_CONFIG_ARCHIVE_MAP,
|
||||
ConditionalDetrConfig,
|
||||
|
@ -5980,6 +5984,7 @@ if TYPE_CHECKING:
|
|||
from .models.clip import CLIPTokenizerFast
|
||||
from .models.code_llama import CodeLlamaTokenizerFast
|
||||
from .models.codegen import CodeGenTokenizerFast
|
||||
from .models.cohere import CohereTokenizerFast
|
||||
from .models.convbert import ConvBertTokenizerFast
|
||||
from .models.cpm import CpmTokenizerFast
|
||||
from .models.deberta import DebertaTokenizerFast
|
||||
|
@ -6557,6 +6562,11 @@ if TYPE_CHECKING:
|
|||
CodeGenModel,
|
||||
CodeGenPreTrainedModel,
|
||||
)
|
||||
from .models.cohere import (
|
||||
CohereForCausalLM,
|
||||
CohereModel,
|
||||
CoherePreTrainedModel,
|
||||
)
|
||||
from .models.conditional_detr import (
|
||||
CONDITIONAL_DETR_PRETRAINED_MODEL_ARCHIVE_LIST,
|
||||
ConditionalDetrForObjectDetection,
|
||||
|
|
|
@ -49,6 +49,7 @@ from . import (
|
|||
clvp,
|
||||
code_llama,
|
||||
codegen,
|
||||
cohere,
|
||||
conditional_detr,
|
||||
convbert,
|
||||
convnext,
|
||||
|
|
|
@ -62,6 +62,7 @@ CONFIG_MAPPING_NAMES = OrderedDict(
|
|||
("clvp", "ClvpConfig"),
|
||||
("code_llama", "LlamaConfig"),
|
||||
("codegen", "CodeGenConfig"),
|
||||
("cohere", "CohereConfig"),
|
||||
("conditional_detr", "ConditionalDetrConfig"),
|
||||
("convbert", "ConvBertConfig"),
|
||||
("convnext", "ConvNextConfig"),
|
||||
|
@ -303,6 +304,7 @@ CONFIG_ARCHIVE_MAP_MAPPING_NAMES = OrderedDict(
|
|||
("clipseg", "CLIPSEG_PRETRAINED_CONFIG_ARCHIVE_MAP"),
|
||||
("clvp", "CLVP_PRETRAINED_CONFIG_ARCHIVE_MAP"),
|
||||
("codegen", "CODEGEN_PRETRAINED_CONFIG_ARCHIVE_MAP"),
|
||||
("cohere", "COHERE_PRETRAINED_CONFIG_ARCHIVE_MAP"),
|
||||
("conditional_detr", "CONDITIONAL_DETR_PRETRAINED_CONFIG_ARCHIVE_MAP"),
|
||||
("convbert", "CONVBERT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
|
||||
("convnext", "CONVNEXT_PRETRAINED_CONFIG_ARCHIVE_MAP"),
|
||||
|
@ -533,6 +535,7 @@ MODEL_NAMES_MAPPING = OrderedDict(
|
|||
("clvp", "CLVP"),
|
||||
("code_llama", "CodeLlama"),
|
||||
("codegen", "CodeGen"),
|
||||
("cohere", "Cohere"),
|
||||
("conditional_detr", "Conditional DETR"),
|
||||
("convbert", "ConvBERT"),
|
||||
("convnext", "ConvNeXT"),
|
||||
|
|
|
@ -64,6 +64,7 @@ MODEL_MAPPING_NAMES = OrderedDict(
|
|||
("clvp", "ClvpModelForConditionalGeneration"),
|
||||
("code_llama", "LlamaModel"),
|
||||
("codegen", "CodeGenModel"),
|
||||
("cohere", "CohereModel"),
|
||||
("conditional_detr", "ConditionalDetrModel"),
|
||||
("convbert", "ConvBertModel"),
|
||||
("convnext", "ConvNextModel"),
|
||||
|
@ -426,6 +427,7 @@ MODEL_FOR_CAUSAL_LM_MAPPING_NAMES = OrderedDict(
|
|||
("camembert", "CamembertForCausalLM"),
|
||||
("code_llama", "LlamaForCausalLM"),
|
||||
("codegen", "CodeGenForCausalLM"),
|
||||
("cohere", "CohereForCausalLM"),
|
||||
("cpmant", "CpmAntForCausalLM"),
|
||||
("ctrl", "CTRLLMHeadModel"),
|
||||
("data2vec-text", "Data2VecTextForCausalLM"),
|
||||
|
|
|
@ -137,6 +137,7 @@ else:
|
|||
),
|
||||
),
|
||||
("codegen", ("CodeGenTokenizer", "CodeGenTokenizerFast" if is_tokenizers_available() else None)),
|
||||
("cohere", (None, "CohereTokenizerFast" if is_tokenizers_available() else None)),
|
||||
("convbert", ("ConvBertTokenizer", "ConvBertTokenizerFast" if is_tokenizers_available() else None)),
|
||||
(
|
||||
"cpm",
|
||||
|
|
|
@ -0,0 +1,77 @@
|
|||
# Copyright 2024 Cohere and The HuggingFace Inc. team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
from typing import TYPE_CHECKING
|
||||
|
||||
from ...utils import (
|
||||
OptionalDependencyNotAvailable,
|
||||
_LazyModule,
|
||||
is_sentencepiece_available,
|
||||
is_tokenizers_available,
|
||||
is_torch_available,
|
||||
)
|
||||
|
||||
|
||||
_import_structure = {
|
||||
"configuration_cohere": ["COHERE_PRETRAINED_CONFIG_ARCHIVE_MAP", "CohereConfig"],
|
||||
}
|
||||
|
||||
|
||||
try:
|
||||
if not is_tokenizers_available():
|
||||
raise OptionalDependencyNotAvailable()
|
||||
except OptionalDependencyNotAvailable:
|
||||
pass
|
||||
else:
|
||||
_import_structure["tokenization_cohere_fast"] = ["CohereTokenizerFast"]
|
||||
|
||||
try:
|
||||
if not is_torch_available():
|
||||
raise OptionalDependencyNotAvailable()
|
||||
except OptionalDependencyNotAvailable:
|
||||
pass
|
||||
else:
|
||||
_import_structure["modeling_cohere"] = [
|
||||
"CohereForCausalLM",
|
||||
"CohereModel",
|
||||
"CoherePreTrainedModel",
|
||||
]
|
||||
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from .configuration_cohere import COHERE_PRETRAINED_CONFIG_ARCHIVE_MAP, CohereConfig
|
||||
|
||||
try:
|
||||
if not is_tokenizers_available():
|
||||
raise OptionalDependencyNotAvailable()
|
||||
except OptionalDependencyNotAvailable:
|
||||
pass
|
||||
else:
|
||||
from .tokenization_cohere_fast import CohereTokenizerFast
|
||||
|
||||
try:
|
||||
if not is_torch_available():
|
||||
raise OptionalDependencyNotAvailable()
|
||||
except OptionalDependencyNotAvailable:
|
||||
pass
|
||||
else:
|
||||
from .modeling_cohere import (
|
||||
CohereForCausalLM,
|
||||
CohereModel,
|
||||
CoherePreTrainedModel,
|
||||
)
|
||||
|
||||
else:
|
||||
import sys
|
||||
|
||||
sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__)
|
|
@ -0,0 +1,155 @@
|
|||
# coding=utf-8
|
||||
# Copyright 2024 Cohere team. All rights reserved.
|
||||
#
|
||||
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
|
||||
# and OPT implementations in this library. It has been modified from its
|
||||
# original forms to accommodate minor architectural differences compared
|
||||
# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
""" Cohere model configuration"""
|
||||
|
||||
from ...configuration_utils import PretrainedConfig
|
||||
from ...utils import logging
|
||||
|
||||
|
||||
logger = logging.get_logger(__name__)
|
||||
|
||||
COHERE_PRETRAINED_CONFIG_ARCHIVE_MAP = {}
|
||||
|
||||
|
||||
class CohereConfig(PretrainedConfig):
|
||||
r"""
|
||||
This is the configuration class to store the configuration of a [`CohereModel`]. It is used to instantiate an Cohere
|
||||
model according to the specified arguments, defining the model architecture.
|
||||
|
||||
Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
|
||||
documentation from [`PretrainedConfig`] for more information. Instantiating a configuration
|
||||
with the defaults will yield a similar configuration to that of the [CohereForAI/c4ai-command-r-v01](https://huggingface.co/CohereForAI/c4ai-command-r-v01) model.
|
||||
|
||||
|
||||
Args:
|
||||
vocab_size (`int`, *optional*, defaults to 256000):
|
||||
Vocabulary size of the Cohere model. Defines the number of different tokens that can be represented by the
|
||||
`inputs_ids` passed when calling [`CohereModel`]
|
||||
hidden_size (`int`, *optional*, defaults to 8192):
|
||||
Dimension of the hidden representations.
|
||||
intermediate_size (`int`, *optional*, defaults to 22528):
|
||||
Dimension of the MLP representations.
|
||||
logit_scale (`float`, *optional*, defaults to 0.0625):
|
||||
The scaling factor for the output logits.
|
||||
num_hidden_layers (`int`, *optional*, defaults to 40):
|
||||
Number of hidden layers in the Transformer decoder.
|
||||
num_attention_heads (`int`, *optional*, defaults to 64):
|
||||
Number of attention heads for each attention layer in the Transformer decoder.
|
||||
num_key_value_heads (`int`, *optional*):
|
||||
This is the number of key_value heads that should be used to implement Grouped Query Attention. If
|
||||
`num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
|
||||
`num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. When
|
||||
converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
|
||||
by meanpooling all the original heads within that group. For more details checkout [this
|
||||
paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
|
||||
`num_attention_heads`.
|
||||
hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
|
||||
The non-linear activation function (function or string) in the decoder.
|
||||
max_position_embeddings (`int`, *optional*, defaults to 8192):
|
||||
The maximum sequence length that this model might ever be used with.
|
||||
initializer_range (`float`, *optional*, defaults to 0.02):
|
||||
The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
|
||||
layer_norm_eps (`float`, *optional*, defaults to 1e-05):
|
||||
The epsilon used by the layer normalization.
|
||||
use_cache (`bool`, *optional*, defaults to `True`):
|
||||
Whether or not the model should return the last key/values attentions (not used by all models). Only
|
||||
relevant if `config.is_decoder=True`.
|
||||
pad_token_id (`int`, *optional*, defaults to 0):
|
||||
Padding token id.
|
||||
bos_token_id (`int`, *optional*, defaults to 5):
|
||||
Beginning of stream token id.
|
||||
eos_token_id (`int`, *optional*, defaults to 255001):
|
||||
End of stream token id.
|
||||
tie_word_embeddings (`bool`, *optional*, defaults to `True`):
|
||||
Whether to tie weight embeddings
|
||||
rope_theta (`float`, *optional*, defaults to 10000.0):
|
||||
The base period of the RoPE embeddings.
|
||||
attention_bias (`bool`, defaults to `False`, *optional*, defaults to `False`):
|
||||
Whether to use a bias in the query, key, value and output projection layers during self-attention.
|
||||
attention_dropout (`float`, *optional*, defaults to 0.0):
|
||||
The dropout ratio for the attention probabilities.
|
||||
|
||||
```python
|
||||
>>> from transformers import CohereModel, CohereConfig
|
||||
|
||||
>>> # Initializing a Cohere model configuration
|
||||
>>> configuration = CohereConfig()
|
||||
|
||||
>>> # Initializing a model from the Cohere configuration
|
||||
>>> model = CohereModel(configuration) # doctest: +SKIP
|
||||
|
||||
>>> # Accessing the model configuration
|
||||
>>> configuration = model.config # doctest: +SKIP
|
||||
```"""
|
||||
|
||||
model_type = "cohere"
|
||||
keys_to_ignore_at_inference = ["past_key_values"]
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
vocab_size=256000,
|
||||
hidden_size=8192,
|
||||
intermediate_size=22528,
|
||||
logit_scale=0.0625,
|
||||
num_hidden_layers=40,
|
||||
num_attention_heads=64,
|
||||
num_key_value_heads=None,
|
||||
hidden_act="silu",
|
||||
max_position_embeddings=8192,
|
||||
initializer_range=0.02,
|
||||
layer_norm_eps=1e-5,
|
||||
use_cache=True,
|
||||
pad_token_id=0,
|
||||
bos_token_id=5,
|
||||
eos_token_id=255001,
|
||||
tie_word_embeddings=True,
|
||||
rope_theta=10000.0,
|
||||
attention_bias=False,
|
||||
attention_dropout=0.0,
|
||||
**kwargs,
|
||||
):
|
||||
self.vocab_size = vocab_size
|
||||
self.max_position_embeddings = max_position_embeddings
|
||||
self.hidden_size = hidden_size
|
||||
self.logit_scale = logit_scale
|
||||
self.intermediate_size = intermediate_size
|
||||
self.num_hidden_layers = num_hidden_layers
|
||||
self.num_attention_heads = num_attention_heads
|
||||
|
||||
# for backward compatibility
|
||||
if num_key_value_heads is None:
|
||||
num_key_value_heads = num_attention_heads
|
||||
|
||||
self.num_key_value_heads = num_key_value_heads
|
||||
self.hidden_act = hidden_act
|
||||
self.initializer_range = initializer_range
|
||||
self.layer_norm_eps = layer_norm_eps
|
||||
self.use_cache = use_cache
|
||||
self.rope_theta = rope_theta
|
||||
self.attention_bias = attention_bias
|
||||
self.attention_dropout = attention_dropout
|
||||
|
||||
super().__init__(
|
||||
pad_token_id=pad_token_id,
|
||||
bos_token_id=bos_token_id,
|
||||
eos_token_id=eos_token_id,
|
||||
tie_word_embeddings=tie_word_embeddings,
|
||||
**kwargs,
|
||||
)
|
File diff suppressed because it is too large
Load Diff
|
@ -0,0 +1,701 @@
|
|||
# coding=utf-8
|
||||
# Copyright 2024 Cohere team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
# This file is based on the tokenization_llama_fast.py file in transformers
|
||||
|
||||
import pickle
|
||||
from typing import Dict, List, Literal, Union
|
||||
|
||||
from tokenizers import processors
|
||||
|
||||
from ...pipelines.conversational import Conversation
|
||||
from ...tokenization_utils_base import BatchEncoding
|
||||
from ...tokenization_utils_fast import PreTrainedTokenizerFast
|
||||
from ...utils import logging
|
||||
from ...utils.versions import require_version
|
||||
|
||||
|
||||
require_version("tokenizers>=0.13.3")
|
||||
|
||||
logger = logging.get_logger(__name__)
|
||||
VOCAB_FILES_NAMES = {"tokenizer_file": "tokenizer.json"}
|
||||
|
||||
PRETRAINED_VOCAB_FILES_MAP = {
|
||||
"tokenizer_file": {
|
||||
"Cohere/Command-nightly": "https://huggingface.co/Cohere/Command-nightly/blob/main/tokenizer.json",
|
||||
},
|
||||
}
|
||||
|
||||
# fmt: off
|
||||
DEFAULT_SYSTEM_PROMPT = "You are Command-R, a brilliant, sophisticated, AI-assistant trained to assist human users by providing thorough responses. You are trained by Cohere."
|
||||
DEFAULT_RAG_PREAMBLE = """## Task and Context
|
||||
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
|
||||
|
||||
## Style Guide
|
||||
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling."""
|
||||
# fmt: on
|
||||
|
||||
|
||||
class CohereTokenizerFast(PreTrainedTokenizerFast):
|
||||
"""
|
||||
Construct a Cohere tokenizer. Based on byte-level Byte-Pair-Encoding.
|
||||
|
||||
This uses notably ByteFallback and NFC normalization.
|
||||
|
||||
```python
|
||||
>>> from transformers import AutoTokenizer
|
||||
|
||||
>>> tokenizer = AutoTokenizer.from_pretrained("CohereForAI/c4ai-command-r-v01")
|
||||
>>> tokenizer.encode("Hello this is a test")
|
||||
[5, 28339, 2075, 1801, 1671, 3282]
|
||||
```
|
||||
|
||||
If you want to change the `bos_token` or the `eos_token`, make sure to specify them when initializing the model, or
|
||||
call `tokenizer.update_post_processor()` to make sure that the post-processing is correctly done (otherwise the
|
||||
values of the first token and final token of an encoded sequence will not be correct). For more details, checkout
|
||||
[post-processors] (https://huggingface.co/docs/tokenizers/api/post-processors) documentation.
|
||||
|
||||
You can get around that behavior by passing `add_prefix_space=True` when instantiating this tokenizer, but since
|
||||
the model was not pretrained this way, it might yield a decrease in performance.
|
||||
|
||||
<Tip>
|
||||
|
||||
When used with `is_split_into_words=True`, this tokenizer needs to be instantiated with `add_prefix_space=True`.
|
||||
|
||||
</Tip>
|
||||
|
||||
This tokenizer inherits from [`PreTrainedTokenizerFast`] which contains most of the main methods. Users should
|
||||
refer to this superclass for more information regarding those methods.
|
||||
|
||||
Args:
|
||||
vocab_file (`str`, *optional*):
|
||||
Path to the vocabulary file.
|
||||
merges_file (`str`, *optional*):
|
||||
Path to the merges file.
|
||||
tokenizer_file (`str`, *optional*):
|
||||
[tokenizers](https://github.com/huggingface/tokenizers) file (generally has a .json extension) that
|
||||
contains everything needed to load the tokenizer.
|
||||
clean_up_tokenization_spaces (`bool`, *optional*, defaults to `False`):
|
||||
Whether or not to cleanup spaces after decoding, cleanup consists in removing potential artifacts like
|
||||
extra spaces.
|
||||
unk_token (`str` or `tokenizers.AddedToken`, *optional*, defaults to `"<UNK>"`):
|
||||
The unknown token. A token that is not in the vocabulary cannot be converted to an ID and is set to be this
|
||||
token instead.
|
||||
bos_token (`str` or `tokenizers.AddedToken`, *optional*, defaults to `"<BOS_TOKEN>"`):
|
||||
The beginning of sequence token that was used during pretraining. Can be used a sequence classifier token.
|
||||
eos_token (`str` or `tokenizers.AddedToken`, *optional*, defaults to `"<|END_OF_TURN_TOKEN|>"`):
|
||||
The end of sequence token.
|
||||
add_bos_token (`bool`, *optional*, defaults to `True`):
|
||||
Whether or not to add an `bos_token` at the start of sequences.
|
||||
add_eos_token (`bool`, *optional*, defaults to `False`):
|
||||
Whether or not to add an `eos_token` at the end of sequences.
|
||||
use_default_system_prompt (`bool`, *optional*, defaults to `False`):
|
||||
Whether or not the default system prompt for Cohere tokenizer should be used.
|
||||
add_prefix_space (`bool`, *optional*, defaults to `False`):
|
||||
Whether or not the tokenizer should automatically add a prefix space
|
||||
"""
|
||||
|
||||
vocab_files_names = VOCAB_FILES_NAMES
|
||||
pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
|
||||
padding_side = "left"
|
||||
model_input_names = ["input_ids", "attention_mask"]
|
||||
slow_tokenizer_class = None
|
||||
# No `max_model_input_sizes`
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
vocab_file=None,
|
||||
merges_file=None,
|
||||
tokenizer_file=None,
|
||||
clean_up_tokenization_spaces=False,
|
||||
unk_token="<UNK>",
|
||||
bos_token="<BOS_TOKEN>",
|
||||
eos_token="<|END_OF_TURN_TOKEN|>",
|
||||
add_bos_token=True,
|
||||
add_eos_token=False,
|
||||
use_default_system_prompt=False,
|
||||
add_prefix_space=False,
|
||||
**kwargs,
|
||||
):
|
||||
super().__init__(
|
||||
vocab_file=vocab_file,
|
||||
merges_file=merges_file,
|
||||
tokenizer_file=tokenizer_file,
|
||||
clean_up_tokenization_spaces=clean_up_tokenization_spaces,
|
||||
unk_token=unk_token,
|
||||
bos_token=bos_token,
|
||||
eos_token=eos_token,
|
||||
add_bos_token=add_bos_token,
|
||||
add_eos_token=add_eos_token,
|
||||
use_default_system_prompt=use_default_system_prompt,
|
||||
add_prefix_space=add_prefix_space,
|
||||
**kwargs,
|
||||
)
|
||||
self._add_bos_token = add_bos_token
|
||||
self._add_eos_token = add_eos_token
|
||||
self.update_post_processor()
|
||||
self.use_default_system_prompt = use_default_system_prompt
|
||||
self.vocab_file = vocab_file
|
||||
self.grounded_generation_template = kwargs.pop("grounded_generation_template", None)
|
||||
self.tool_use_template = kwargs.pop("tool_use_template", None)
|
||||
|
||||
# TODO @ArthurZucker this can only work one way for now, to update later-on. Tests should also properly
|
||||
# check this as they were green before.
|
||||
pre_tok_state = pickle.dumps(self.backend_tokenizer.pre_tokenizer)
|
||||
decoder_state = pickle.dumps(self.backend_tokenizer.decoder)
|
||||
|
||||
if add_prefix_space:
|
||||
pre_tok_state = pre_tok_state.replace(b'"add_prefix_space":false', b'"add_prefix_space": true')
|
||||
decoder_state = decoder_state.replace(b'"add_prefix_space":false', b'"add_prefix_space": true')
|
||||
self.backend_tokenizer.pre_tokenizer = pickle.loads(pre_tok_state)
|
||||
self.backend_tokenizer.decoder = pickle.loads(decoder_state)
|
||||
|
||||
self.add_prefix_space = add_prefix_space
|
||||
|
||||
def _batch_encode_plus(self, *args, **kwargs) -> BatchEncoding:
|
||||
is_split_into_words = kwargs.get("is_split_into_words", False)
|
||||
if not (self.add_prefix_space or not is_split_into_words):
|
||||
raise Exception(
|
||||
f"You need to instantiate {self.__class__.__name__} with add_prefix_space=True to use it with"
|
||||
" pretokenized inputs."
|
||||
)
|
||||
|
||||
return super()._batch_encode_plus(*args, **kwargs)
|
||||
|
||||
def _encode_plus(self, *args, **kwargs) -> BatchEncoding:
|
||||
is_split_into_words = kwargs.get("is_split_into_words", False)
|
||||
|
||||
if not (self.add_prefix_space or not is_split_into_words):
|
||||
raise Exception(
|
||||
f"You need to instantiate {self.__class__.__name__} with add_prefix_space=True to use it with"
|
||||
" pretokenized inputs."
|
||||
)
|
||||
|
||||
return super()._encode_plus(*args, **kwargs)
|
||||
|
||||
def update_post_processor(self):
|
||||
"""
|
||||
Updates the underlying post processor with the current `bos_token` and `eos_token`.
|
||||
"""
|
||||
bos = self.bos_token
|
||||
bos_token_id = self.bos_token_id
|
||||
if bos is None and self.add_bos_token:
|
||||
raise ValueError("add_bos_token = True but bos_token = None")
|
||||
|
||||
eos = self.eos_token
|
||||
eos_token_id = self.eos_token_id
|
||||
if eos is None and self.add_eos_token:
|
||||
raise ValueError("add_eos_token = True but eos_token = None")
|
||||
|
||||
single = f"{(bos+':0 ') if self.add_bos_token else ''}$A:0{(' '+eos+':0') if self.add_eos_token else ''}"
|
||||
pair = f"{single}{(' '+bos+':1') if self.add_bos_token else ''} $B:1{(' '+eos+':1') if self.add_eos_token else ''}"
|
||||
|
||||
special_tokens = []
|
||||
if self.add_bos_token:
|
||||
special_tokens.append((bos, bos_token_id))
|
||||
if self.add_eos_token:
|
||||
special_tokens.append((eos, eos_token_id))
|
||||
self._tokenizer.post_processor = processors.TemplateProcessing(
|
||||
single=single, pair=pair, special_tokens=special_tokens
|
||||
)
|
||||
|
||||
@property
|
||||
def add_eos_token(self):
|
||||
return self._add_eos_token
|
||||
|
||||
@property
|
||||
def add_bos_token(self):
|
||||
return self._add_bos_token
|
||||
|
||||
@add_eos_token.setter
|
||||
def add_eos_token(self, value):
|
||||
self._add_eos_token = value
|
||||
self.update_post_processor()
|
||||
|
||||
@add_bos_token.setter
|
||||
def add_bos_token(self, value):
|
||||
self._add_bos_token = value
|
||||
self.update_post_processor()
|
||||
|
||||
@property
|
||||
def default_chat_template(self):
|
||||
"""
|
||||
Cohere Tokenizer uses <|START_OF_TURN_TOKEN|> and <|END_OF_TURN_TOKEN|> to indicate each turn in a chat.
|
||||
Additioanlly, to indicate the source of the message, <|USER_TOKEN|>, <|CHATBOT_TOKEN|> and <|SYSTEM_TOKEN|>
|
||||
for user, assitant and system messages respectively.
|
||||
|
||||
The output should look something like:
|
||||
<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>{{ preamble }}<|END_OF_TURN_TOKEN|><BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>{{ How are you? }}<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>{{ I am doing well! }}<|END_OF_TURN_TOKEN|>
|
||||
|
||||
Use add_generation_prompt to add a prompt for the model to generate a response:
|
||||
>>> from transformers import AutoTokenizer
|
||||
>>> tokenizer = AutoTokenizer.from_pretrained("CohereForAI/c4ai-command-r-v01")
|
||||
>>> messages = [{"role": "user", "content": "Hello, how are you?"}]
|
||||
>>> tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
||||
'<BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Hello, how are you?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>'
|
||||
|
||||
"""
|
||||
logger.warning_once(
|
||||
"\nNo chat template is defined for this tokenizer - using the default template "
|
||||
f"for the {self.__class__.__name__} class. If the default is not appropriate for "
|
||||
"your model, please set `tokenizer.chat_template` to an appropriate template. "
|
||||
"See https://huggingface.co/docs/transformers/main/chat_templating for more information.\n"
|
||||
)
|
||||
default_template = (
|
||||
"{{ bos_token }}"
|
||||
"{% if messages[0]['role'] == 'system' %}"
|
||||
"{% set loop_messages = messages[1:] %}" # Extract system message if it's present
|
||||
"{% set system_message = messages[0]['content'] %}"
|
||||
"{% elif USE_DEFAULT_PROMPT == true %}"
|
||||
"{% set loop_messages = messages %}" # Or use the default system message if the flag is set
|
||||
"{% set system_message = 'DEFAULT_SYSTEM_MESSAGE' %}"
|
||||
"{% else %}"
|
||||
"{% set loop_messages = messages %}"
|
||||
"{% set system_message = false %}"
|
||||
"{% endif %}"
|
||||
"{% if system_message != false %}" # Start with system message
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' + system_message + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% endif %}"
|
||||
"{% for message in loop_messages %}" # Loop over all non-system messages
|
||||
"{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}"
|
||||
"{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}"
|
||||
"{% endif %}"
|
||||
"{% set content = message['content'] %}"
|
||||
"{% if message['role'] == 'user' %}" # After all of that, handle messages/roles in a fairly normal way
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|USER_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% elif message['role'] == 'assistant' %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% endif %}"
|
||||
"{% endfor %}"
|
||||
"{% if add_generation_prompt %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' }}"
|
||||
"{% endif %}"
|
||||
)
|
||||
default_template = default_template.replace(
|
||||
"USE_DEFAULT_PROMPT", "true" if self.use_default_system_prompt else "false"
|
||||
)
|
||||
default_message = DEFAULT_SYSTEM_PROMPT.replace("\n", "\\n").replace("'", "\\'")
|
||||
default_template = default_template.replace("DEFAULT_SYSTEM_MESSAGE", default_message)
|
||||
|
||||
tool_use_template = (
|
||||
"{{ bos_token }}"
|
||||
"{% if messages[0]['role'] == 'system' %}"
|
||||
"{% set loop_messages = messages[1:] %}" # Extract system message if it's present
|
||||
"{% set system_message = messages[0]['content'] %}"
|
||||
"{% else %}"
|
||||
"{% set loop_messages = messages %}"
|
||||
"{% set system_message = 'DEFAULT_SYSTEM_MESSAGE' %}"
|
||||
"{% endif %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' }}"
|
||||
"{{ '# Safety Preamble' }}"
|
||||
"{{ '\nThe instructions in this section override those in the task description and style guide sections. Don\\'t answer questions that are harmful or immoral.' }}"
|
||||
"{{ '\n\n# System Preamble' }}"
|
||||
"{{ '\n## Basic Rules' }}"
|
||||
"{{ '\nYou are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user\\'s requests, you cite your sources in your answers, according to those instructions.' }}"
|
||||
"{{ '\n\n# User Preamble' }}"
|
||||
"{{ '\n' + system_message }}"
|
||||
"{{'\n\n## Available Tools\nHere is a list of tools that you have available to you:\n\n'}}"
|
||||
"{% for tool in tools %}"
|
||||
"{% if loop.index0 != 0 %}"
|
||||
"{{ '\n\n'}}"
|
||||
"{% endif %}"
|
||||
"{{'```python\ndef ' + tool.name + '('}}"
|
||||
"{% for param_name, param_fields in tool.parameter_definitions.items() %}"
|
||||
"{% if loop.index0 != 0 %}"
|
||||
"{{ ', '}}"
|
||||
"{% endif %}"
|
||||
"{{param_name}}: "
|
||||
"{% if not param_fields.required %}"
|
||||
"{{'Optional[' + param_fields.type + '] = None'}}"
|
||||
"{% else %}"
|
||||
"{{ param_fields.type }}"
|
||||
"{% endif %}"
|
||||
"{% endfor %}"
|
||||
'{{ \') -> List[Dict]:\n """\'}}'
|
||||
"{{ tool.description }}"
|
||||
"{% if tool.parameter_definitions|length != 0 %}"
|
||||
"{{ '\n\n Args:\n '}}"
|
||||
"{% for param_name, param_fields in tool.parameter_definitions.items() %}"
|
||||
"{% if loop.index0 != 0 %}"
|
||||
"{{ '\n ' }}"
|
||||
"{% endif %}"
|
||||
"{{ param_name + ' ('}}"
|
||||
"{% if not param_fields.required %}"
|
||||
"{{'Optional[' + param_fields.type + ']'}}"
|
||||
"{% else %}"
|
||||
"{{ param_fields.type }}"
|
||||
"{% endif %}"
|
||||
"{{ '): ' + param_fields.description }}"
|
||||
"{% endfor %}"
|
||||
"{% endif %}"
|
||||
'{{ \'\n """\n pass\n```\' }}'
|
||||
"{% endfor %}"
|
||||
"{{ '<|END_OF_TURN_TOKEN|>'}}"
|
||||
"{% for message in loop_messages %}"
|
||||
"{% set content = message['content'] %}"
|
||||
"{% if message['role'] == 'user' %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|USER_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% elif message['role'] == 'system' %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% elif message['role'] == 'assistant' %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% endif %}"
|
||||
"{% endfor %}"
|
||||
"{{'<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write \\'Action:\\' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user\\'s last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:\n```json\n[\n {\n \"tool_name\": title of the tool in the specification,\n \"parameters\": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters\n }\n]```<|END_OF_TURN_TOKEN|>'}}"
|
||||
"{% if add_generation_prompt %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' }}"
|
||||
"{% endif %}"
|
||||
)
|
||||
default_tool_message = DEFAULT_RAG_PREAMBLE.replace("\n", "\\n").replace("'", "\\'")
|
||||
tool_use_template = tool_use_template.replace("DEFAULT_SYSTEM_MESSAGE", default_tool_message)
|
||||
|
||||
rag_template = (
|
||||
"{{ bos_token }}"
|
||||
"{% if messages[0]['role'] == 'system' %}"
|
||||
"{% set loop_messages = messages[1:] %}" # Extract system message if it's present
|
||||
"{% set system_message = messages[0]['content'] %}"
|
||||
"{% else %}"
|
||||
"{% set loop_messages = messages %}"
|
||||
"{% set system_message = 'DEFAULT_SYSTEM_MESSAGE' %}"
|
||||
"{% endif %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' }}"
|
||||
"{{ '# Safety Preamble' }}"
|
||||
"{{ '\nThe instructions in this section override those in the task description and style guide sections. Don\\'t answer questions that are harmful or immoral.' }}"
|
||||
"{{ '\n\n# System Preamble' }}"
|
||||
"{{ '\n## Basic Rules' }}"
|
||||
"{{ '\nYou are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user\\'s requests, you cite your sources in your answers, according to those instructions.' }}"
|
||||
"{{ '\n\n# User Preamble' }}"
|
||||
"{{ '\n' + system_message }}"
|
||||
"{{ '<|END_OF_TURN_TOKEN|>'}}"
|
||||
"{% for message in loop_messages %}" # Loop over all non-system messages
|
||||
"{% set content = message['content'] %}"
|
||||
"{% if message['role'] == 'user' %}" # After all of that, handle messages/roles in a fairly normal way
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|USER_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% elif message['role'] == 'system' %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% elif message['role'] == 'assistant' %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' + content.strip() + '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% endif %}"
|
||||
"{% endfor %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>'}}"
|
||||
"{{ '<results>' }}"
|
||||
"{% for document in documents %}" # Loop over all non-system messages
|
||||
"{{ '\nDocument: ' }}"
|
||||
"{{ loop.index0 }}\n"
|
||||
"{% for key, value in document.items() %}"
|
||||
"{{ key }}: {{value}}\n"
|
||||
"{% endfor %}"
|
||||
"{% endfor %}"
|
||||
"{{ '</results>'}}"
|
||||
"{{ '<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>' }}"
|
||||
"{{ 'Carefully perform the following instructions, in order, starting each with a new line.\n' }}"
|
||||
"{{ 'Firstly, Decide which of the retrieved documents are relevant to the user\\'s last input by writing \\'Relevant Documents:\\' followed by comma-separated list of document numbers. If none are relevant, you should instead write \\'None\\'.\n' }}"
|
||||
"{{ 'Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user\\'s last input by writing \\'Cited Documents:\\' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write \\'None\\'.\n' }}"
|
||||
"{% if citation_mode=='accurate' %}"
|
||||
"{{ 'Thirdly, Write \\'Answer:\\' followed by a response to the user\\'s last input in high quality natural english. Use the retrieved documents to help you. Do not insert any citations or grounding markup.\n' }}"
|
||||
"{% endif %}"
|
||||
"{{ 'Finally, Write \\'Grounded answer:\\' followed by a response to the user\\'s last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.' }}"
|
||||
"{{ '<|END_OF_TURN_TOKEN|>' }}"
|
||||
"{% if add_generation_prompt %}"
|
||||
"{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' }}"
|
||||
"{% endif %}"
|
||||
)
|
||||
default_rag_message = DEFAULT_RAG_PREAMBLE.replace("\n", "\\n").replace("'", "\\'")
|
||||
rag_template = rag_template.replace("DEFAULT_SYSTEM_MESSAGE", default_rag_message)
|
||||
|
||||
return {"default": default_template, "tool_use": tool_use_template, "rag": rag_template}
|
||||
|
||||
def apply_tool_use_template(
|
||||
self,
|
||||
conversation: Union[List[Dict[str, str]], "Conversation"],
|
||||
tools: List[Dict],
|
||||
**kwargs,
|
||||
) -> Union[str, List[int]]:
|
||||
"""Create a Command-R tool-use prompt.
|
||||
|
||||
Once rendered, the prompt instructs the model to generate a list of actions to perform on a set of user supplied tools
|
||||
to help carry out the user's requests.
|
||||
|
||||
Conceptually, this works in the same way as `apply_chat_format`, but takes an additional `tools` parameter.
|
||||
|
||||
Converts a Conversation object or a list of dictionaries with `"role"` and `"content"` keys and a list of available
|
||||
tools for the model to use into a prompt string, or a list of token ids.
|
||||
This method will use the tokenizer's `default_tool_use_template` template specified at the class level.
|
||||
You can override the default template using the `tool_use_template` kwarg but the quality of your results may decrease.
|
||||
|
||||
Args:
|
||||
conversation (Union[List[Dict[str, str]], "Conversation"]): A Conversation object or list of dicts
|
||||
with "role" and "content" keys, representing the chat history so far.
|
||||
tools (List[Dict]): a list of tools to render into the prompt for the model to choose from.
|
||||
See an example at the bottom of the docstring.
|
||||
The format should be:
|
||||
* name (str): The name of the tool to be called. Valid names contain only the characters a-z,
|
||||
A-Z, 0-9, _ and must not begin with a digit.
|
||||
* description (str): The description of what the tool does, the model uses the description to
|
||||
choose when and how to call the function.
|
||||
* parameter_definitions (List[Dict]): The input parameters of the tool. Accepts a dictionary
|
||||
where the key is the name of the parameter and the value is the parameter spec.
|
||||
Valid parameter names contain only the characters a-z, A-Z, 0-9, _ and must not begin with a digit.
|
||||
Parameter specs are as follows:
|
||||
* description (str): The description of the parameter.
|
||||
* type (str): the type of the parameter - most effective for python builtin data types, such as 'str', 'bool'
|
||||
* required: boolean: Denotes whether the parameter is always present (required) or not. Defaults to not required.
|
||||
add_generation_prompt (bool, *optional*): Whether to end the prompt with the token(s) that indicate
|
||||
the start of an assistant message. This is useful when you want to generate a response from the model.
|
||||
Note that this argument will be passed to the chat template, and so it must be supported in the
|
||||
template for this argument to have any effect.
|
||||
tokenize (`bool`, defaults to `True`):
|
||||
Whether to tokenize the output. If `False`, the output will be a string.
|
||||
padding (`bool`, defaults to `False`):
|
||||
Whether to pad sequences to the maximum length. Has no effect if tokenize is `False`.
|
||||
truncation (`bool`, defaults to `False`):
|
||||
Whether to truncate sequences at the maximum length. Has no effect if tokenize is `False`.
|
||||
max_length (`int`, *optional*):
|
||||
Maximum length (in tokens) to use for padding or truncation. Has no effect if tokenize is `False`. If
|
||||
not specified, the tokenizer's `max_length` attribute will be used as a default.
|
||||
return_tensors (`str` or [`~utils.TensorType`], *optional*):
|
||||
If set, will return tensors of a particular framework. Has no effect if tokenize is `False`. Acceptable
|
||||
values are:
|
||||
- `'tf'`: Return TensorFlow `tf.Tensor` objects.
|
||||
- `'pt'`: Return PyTorch `torch.Tensor` objects.
|
||||
- `'np'`: Return NumPy `np.ndarray` objects.
|
||||
- `'jax'`: Return JAX `jnp.ndarray` objects.
|
||||
return_dict (`bool`, *optional*, defaults to `False`):
|
||||
Whether to return a dictionary with named outputs. Has no effect if tokenize is `False`.
|
||||
**tokenizer_kwargs: Additional kwargs to pass to the tokenizer.
|
||||
|
||||
Returns:
|
||||
`str`: A rendered prompt string.
|
||||
or if tokenize=True:
|
||||
`List[int]`: A list of token ids representing the tokenized chat so far, including control tokens. This
|
||||
output is ready to pass to the model, either directly or via methods like `generate()`.
|
||||
|
||||
Examples:
|
||||
|
||||
```python
|
||||
>> tokenizer = CohereTokenizerFast.from_pretrained("CohereForAI/c4ai-command-r-v01")
|
||||
>> tools = [
|
||||
{
|
||||
"name": "internet_search",
|
||||
"description": "Returns a list of relevant document snippets for a textual query retrieved from the internet",
|
||||
"parameter_definitions": {
|
||||
"query": {
|
||||
"description": "Query to search the internet with",
|
||||
"type": "str",
|
||||
"required": True
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"name': "directly_answer",
|
||||
"description": "Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history",
|
||||
"parameter_definitions": {}
|
||||
}
|
||||
]
|
||||
>> conversation = [
|
||||
{"role": "user", "content": "Whats the biggest penguin in the world?"}
|
||||
]
|
||||
>> # render the prompt, ready for user to inspect, or for input into the model:
|
||||
>> prompt = tokenizer.apply_tool_use_template(conversation, tools=tools, tokenize=False, add_generation_prompt=True)
|
||||
>> print(prompt)
|
||||
<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
|
||||
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
|
||||
|
||||
# System Preamble
|
||||
## Basic Rules
|
||||
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
|
||||
|
||||
# User Preamble
|
||||
## Task and Context
|
||||
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
|
||||
|
||||
## Style Guide
|
||||
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.
|
||||
|
||||
## Available Tools
|
||||
Here is a list of tools that you have available to you:
|
||||
|
||||
\\`\\`\\`python
|
||||
def internet_search(query: str) -> List[Dict]:
|
||||
\"\"\"Returns a list of relevant document snippets for a textual query retrieved from the internet
|
||||
|
||||
Args:
|
||||
query (str): Query to search the internet with
|
||||
\"\"\"
|
||||
pass
|
||||
\\`\\`\\`
|
||||
|
||||
\\`\\`\\`python
|
||||
def directly_answer() -> List[Dict]:
|
||||
\"\"\"Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history
|
||||
\"\"\"
|
||||
pass
|
||||
\\`\\`\\`<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
|
||||
\\`\\`\\`json
|
||||
[
|
||||
{
|
||||
"tool_name": title of the tool in the specification,
|
||||
"parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
|
||||
}
|
||||
]\\`\\`\\`<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
|
||||
```
|
||||
>> inputs = tokenizer.encode(prompt, add_special_tokens=False, return_tensors='pt')
|
||||
>> outputs = model.generate(inputs, max_new_tokens=128)
|
||||
>> print(tokenizer.decode(outputs[0]))
|
||||
Action: ```json
|
||||
[
|
||||
{
|
||||
"tool_name": "internet_search",
|
||||
"parameters": {
|
||||
"query": "biggest penguin in the world"
|
||||
}
|
||||
}
|
||||
]
|
||||
```
|
||||
"""
|
||||
return self.apply_chat_template(
|
||||
conversation,
|
||||
chat_template="tool_use",
|
||||
tools=tools,
|
||||
**kwargs,
|
||||
)
|
||||
|
||||
def apply_grounded_generation_template(
|
||||
self,
|
||||
conversation: Union[List[Dict[str, str]], "Conversation"],
|
||||
documents: List[Dict],
|
||||
citation_mode: Literal["fast", "accurate"] = "accurate",
|
||||
**kwargs,
|
||||
) -> Union[str, List[int]]:
|
||||
"""Create a Command-R grounded generation (aka RAG) prompt.
|
||||
|
||||
Once rendered, the prompt instructs the model to generate a response with citations in, based on supplied documents.
|
||||
|
||||
Conceptually, this works in the same way as `apply_chat_format`, but takes additional `documents`
|
||||
and parameter `citation_mode` parameters.
|
||||
|
||||
Converts a Conversation object or a list of dictionaries with `"role"` and `"content"` keys and a list of
|
||||
documents for the model to ground its response on into a prompt string, or a list of token ids.
|
||||
This method will use the tokenizer's `grounded_generation_template` template specified at the class level.
|
||||
You can override the default template using the `grounded_generation_template` kwarg but the quality of your results may decrease.
|
||||
|
||||
Args:
|
||||
conversation (Union[List[Dict[str, str]], "Conversation"]): A Conversation object or list of dicts
|
||||
with "role" and "content" keys, representing the chat history so far.
|
||||
documents (List[Dict[str, str]): A list of dicts, representing documents or tool outputs to ground your
|
||||
generation on. A document is a semistructured dict, wiht a string to string mapping. Common fields are
|
||||
`url`, `title`, `snippet` etc but should be descriptive of the key. They will get rendered into the prompt.
|
||||
citation_mode: either "accurate" (prompt the model to generate an answer first, then rewrite it with citation
|
||||
spans in) or "fast", where the prompt instructs the model to generate an answer with citations in directly.
|
||||
The former has higher quality citations, the latter requires fewer tokens to be generated.
|
||||
add_generation_prompt (bool, *optional*): Whether to end the prompt with the token(s) that indicate
|
||||
the start of an assistant message. This is useful when you want to generate a response from the model.
|
||||
Note that this argument will be passed to the chat template, and so it must be supported in the
|
||||
template for this argument to have any effect.
|
||||
tokenize (`bool`, defaults to `True`):
|
||||
Whether to tokenize the output. If `False`, the output will be a string.
|
||||
padding (`bool`, defaults to `False`):
|
||||
Whether to pad sequences to the maximum length. Has no effect if tokenize is `False`.
|
||||
truncation (`bool`, defaults to `False`):
|
||||
Whether to truncate sequences at the maximum length. Has no effect if tokenize is `False`.
|
||||
max_length (`int`, *optional*):
|
||||
Maximum length (in tokens) to use for padding or truncation. Has no effect if tokenize is `False`. If
|
||||
not specified, the tokenizer's `max_length` attribute will be used as a default.
|
||||
return_tensors (`str` or [`~utils.TensorType`], *optional*):
|
||||
If set, will return tensors of a particular framework. Has no effect if tokenize is `False`. Acceptable
|
||||
values are:
|
||||
- `'tf'`: Return TensorFlow `tf.Tensor` objects.
|
||||
- `'pt'`: Return PyTorch `torch.Tensor` objects.
|
||||
- `'np'`: Return NumPy `np.ndarray` objects.
|
||||
- `'jax'`: Return JAX `jnp.ndarray` objects.
|
||||
return_dict (`bool`, *optional*, defaults to `False`):
|
||||
Whether to return a dictionary with named outputs. Has no effect if tokenize is `False`.
|
||||
**tokenizer_kwargs: Additional kwargs to pass to the tokenizer.
|
||||
|
||||
Returns:
|
||||
`str`: A rendered prompt string.
|
||||
or if tokenize=True:
|
||||
`List[int]`: A list of token ids representing the tokenized chat so far, including control tokens. This
|
||||
output is ready to pass to the model, either directly or via methods like `generate()`.
|
||||
|
||||
Examples:
|
||||
|
||||
```python
|
||||
>> tokenizer = CohereTokenizerFast.from_pretrained('CohereForAI/c4ai-command-r-v01')
|
||||
|
||||
>> # define documents:
|
||||
>> documents = [
|
||||
{ "title": "Tall penguins", "text": "Emperor penguins are the tallest." },
|
||||
{ "title": "Penguin habitats", "text": "Emperor penguins only live in Antarctica."}
|
||||
]
|
||||
>> # define a conversation:
|
||||
>> conversation = [
|
||||
{"role": "user", "content": "Whats the biggest penguin in the world?"}
|
||||
]
|
||||
>> # render the prompt, ready for user to inspect, or for input into the model:
|
||||
>> grounded_generation_prompt = tokenizer.apply_grounded_generation_template(conversation, documents=documents, tokenize=False, add_generation_prompt=True)
|
||||
>> print(grounded_generation_prompt)
|
||||
<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
|
||||
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
|
||||
|
||||
## Basic Rules
|
||||
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
|
||||
|
||||
# User Preamble
|
||||
## Task and Context
|
||||
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
|
||||
|
||||
## Style Guide
|
||||
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
|
||||
Document: 0
|
||||
title: Tall penguins
|
||||
text: Emperor penguins are the tallest.
|
||||
|
||||
Document: 1
|
||||
title: Penguin habitats
|
||||
text: Emperor penguins only live in Antarctica.
|
||||
</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Carefully perform the following instructions, in order, starting each with a new line.
|
||||
Firstly, Decide which of the retrieved documents are relevant to the user's last input by writing 'Relevant Documents:' followed by comma-separated list of document numbers. If none are relevant, you should instead write 'None'.
|
||||
Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user's last input by writing 'Cited Documents:' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write 'None'.
|
||||
Thirdly, Write 'Answer:' followed by a response to the user's last input in high quality natural english. Use the retrieved documents to help you. Do not insert any citations or grounding markup.
|
||||
Finally, Write 'Grounded answer:' followed by a response to the user's last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>'''
|
||||
```
|
||||
>> inputs = tokenizer.encode(prompt, add_special_tokens=False, return_tensors='pt')
|
||||
>> outputs = model.generate(inputs, max_new_tokens=128)
|
||||
>> print(tokenizer.decode(outputs[0]))
|
||||
Relevant Documents: 0,1
|
||||
Cited Documents: 0,1
|
||||
Answer: The Emperor Penguin is the tallest or biggest penguin in the world. It is a bird that lives only in Antarctica and grows to a height of around 122 centimetres.
|
||||
Grounded answer: The <co: 0>Emperor Penguin</co: 0> is the <co: 0>tallest</co: 0> or biggest penguin in the world. It is a bird that <co: 1>lives only in Antarctica</co: 1> and <co: 0>grows to a height of around 122 centimetres.</co: 0>
|
||||
"""
|
||||
return self.apply_chat_template(
|
||||
conversation,
|
||||
chat_template="rag",
|
||||
documents=documents,
|
||||
citation_mode=citation_mode,
|
||||
**kwargs,
|
||||
)
|
||||
|
||||
# TODO ArthurZ let's rely on the template processor instead, refactor all fast tokenizers
|
||||
def build_inputs_with_special_tokens(self, token_ids_0, token_ids_1=None):
|
||||
bos_token_id = [self.bos_token_id] if self.add_bos_token else []
|
||||
eos_token_id = [self.eos_token_id] if self.add_eos_token else []
|
||||
|
||||
output = bos_token_id + token_ids_0 + eos_token_id
|
||||
|
||||
if token_ids_1 is not None:
|
||||
output = output + bos_token_id + token_ids_1 + eos_token_id
|
||||
|
||||
return output
|
|
@ -2056,6 +2056,27 @@ class CodeGenPreTrainedModel(metaclass=DummyObject):
|
|||
requires_backends(self, ["torch"])
|
||||
|
||||
|
||||
class CohereForCausalLM(metaclass=DummyObject):
|
||||
_backends = ["torch"]
|
||||
|
||||
def __init__(self, *args, **kwargs):
|
||||
requires_backends(self, ["torch"])
|
||||
|
||||
|
||||
class CohereModel(metaclass=DummyObject):
|
||||
_backends = ["torch"]
|
||||
|
||||
def __init__(self, *args, **kwargs):
|
||||
requires_backends(self, ["torch"])
|
||||
|
||||
|
||||
class CoherePreTrainedModel(metaclass=DummyObject):
|
||||
_backends = ["torch"]
|
||||
|
||||
def __init__(self, *args, **kwargs):
|
||||
requires_backends(self, ["torch"])
|
||||
|
||||
|
||||
CONDITIONAL_DETR_PRETRAINED_MODEL_ARCHIVE_LIST = None
|
||||
|
||||
|
||||
|
|
|
@ -86,6 +86,13 @@ class CodeGenTokenizerFast(metaclass=DummyObject):
|
|||
requires_backends(self, ["tokenizers"])
|
||||
|
||||
|
||||
class CohereTokenizerFast(metaclass=DummyObject):
|
||||
_backends = ["tokenizers"]
|
||||
|
||||
def __init__(self, *args, **kwargs):
|
||||
requires_backends(self, ["tokenizers"])
|
||||
|
||||
|
||||
class ConvBertTokenizerFast(metaclass=DummyObject):
|
||||
_backends = ["tokenizers"]
|
||||
|
||||
|
|
|
@ -135,6 +135,7 @@ _REGULAR_SUPPORTED_MODEL_NAMES_AND_TASKS = [
|
|||
"hubert",
|
||||
"layoutlm",
|
||||
"llama",
|
||||
"cohere",
|
||||
"lxmert",
|
||||
"m2m_100",
|
||||
"marian",
|
||||
|
|
|
@ -0,0 +1,422 @@
|
|||
# coding=utf-8
|
||||
# Copyright 2022 The HuggingFace Inc. team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
""" Testing suite for the PyTorch Cohere model. """
|
||||
|
||||
import unittest
|
||||
|
||||
from parameterized import parameterized
|
||||
|
||||
from transformers import CohereConfig, is_torch_available
|
||||
from transformers.testing_utils import (
|
||||
require_bitsandbytes,
|
||||
require_torch,
|
||||
require_torch_multi_gpu,
|
||||
require_torch_sdpa,
|
||||
slow,
|
||||
torch_device,
|
||||
)
|
||||
|
||||
from ...generation.test_utils import GenerationTesterMixin
|
||||
from ...test_configuration_common import ConfigTester
|
||||
from ...test_modeling_common import ModelTesterMixin, ids_tensor
|
||||
from ...test_pipeline_mixin import PipelineTesterMixin
|
||||
|
||||
|
||||
if is_torch_available():
|
||||
import torch
|
||||
|
||||
from transformers import AutoTokenizer, CohereForCausalLM, CohereModel
|
||||
|
||||
|
||||
# Copied from transformers.tests.models.llama.LlamaModelTester with Llama->Cohere
|
||||
class CohereModelTester:
|
||||
def __init__(
|
||||
self,
|
||||
parent,
|
||||
batch_size=13,
|
||||
seq_length=7,
|
||||
is_training=True,
|
||||
use_input_mask=True,
|
||||
use_token_type_ids=False,
|
||||
use_labels=True,
|
||||
vocab_size=99,
|
||||
hidden_size=32,
|
||||
num_hidden_layers=2,
|
||||
num_attention_heads=4,
|
||||
intermediate_size=37,
|
||||
hidden_act="gelu",
|
||||
hidden_dropout_prob=0.1,
|
||||
attention_probs_dropout_prob=0.1,
|
||||
max_position_embeddings=512,
|
||||
type_vocab_size=16,
|
||||
type_sequence_label_size=2,
|
||||
initializer_range=0.02,
|
||||
num_labels=3,
|
||||
num_choices=4,
|
||||
pad_token_id=0,
|
||||
scope=None,
|
||||
):
|
||||
self.parent = parent
|
||||
self.batch_size = batch_size
|
||||
self.seq_length = seq_length
|
||||
self.is_training = is_training
|
||||
self.use_input_mask = use_input_mask
|
||||
self.use_token_type_ids = use_token_type_ids
|
||||
self.use_labels = use_labels
|
||||
self.vocab_size = vocab_size
|
||||
self.hidden_size = hidden_size
|
||||
self.num_hidden_layers = num_hidden_layers
|
||||
self.num_attention_heads = num_attention_heads
|
||||
self.intermediate_size = intermediate_size
|
||||
self.hidden_act = hidden_act
|
||||
self.hidden_dropout_prob = hidden_dropout_prob
|
||||
self.attention_probs_dropout_prob = attention_probs_dropout_prob
|
||||
self.max_position_embeddings = max_position_embeddings
|
||||
self.type_vocab_size = type_vocab_size
|
||||
self.type_sequence_label_size = type_sequence_label_size
|
||||
self.initializer_range = initializer_range
|
||||
self.num_labels = num_labels
|
||||
self.num_choices = num_choices
|
||||
self.pad_token_id = pad_token_id
|
||||
self.scope = scope
|
||||
|
||||
def prepare_config_and_inputs(self):
|
||||
input_ids = ids_tensor([self.batch_size, self.seq_length], self.vocab_size)
|
||||
|
||||
input_mask = None
|
||||
if self.use_input_mask:
|
||||
input_mask = torch.tril(torch.ones(self.batch_size, self.seq_length)).to(torch_device)
|
||||
|
||||
token_type_ids = None
|
||||
if self.use_token_type_ids:
|
||||
token_type_ids = ids_tensor([self.batch_size, self.seq_length], self.type_vocab_size)
|
||||
|
||||
sequence_labels = None
|
||||
token_labels = None
|
||||
choice_labels = None
|
||||
if self.use_labels:
|
||||
sequence_labels = ids_tensor([self.batch_size], self.type_sequence_label_size)
|
||||
token_labels = ids_tensor([self.batch_size, self.seq_length], self.num_labels)
|
||||
choice_labels = ids_tensor([self.batch_size], self.num_choices)
|
||||
|
||||
config = self.get_config()
|
||||
|
||||
return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
|
||||
|
||||
# Ignore copy
|
||||
def get_config(self):
|
||||
return CohereConfig(
|
||||
vocab_size=self.vocab_size,
|
||||
hidden_size=self.hidden_size,
|
||||
num_hidden_layers=self.num_hidden_layers,
|
||||
num_attention_heads=self.num_attention_heads,
|
||||
intermediate_size=self.intermediate_size,
|
||||
hidden_act=self.hidden_act,
|
||||
hidden_dropout_prob=self.hidden_dropout_prob,
|
||||
attention_probs_dropout_prob=self.attention_probs_dropout_prob,
|
||||
max_position_embeddings=self.max_position_embeddings,
|
||||
type_vocab_size=self.type_vocab_size,
|
||||
is_decoder=False,
|
||||
initializer_range=self.initializer_range,
|
||||
pad_token_id=self.pad_token_id,
|
||||
eos_token_id=self.pad_token_id,
|
||||
)
|
||||
|
||||
def create_and_check_model(
|
||||
self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
|
||||
):
|
||||
model = CohereModel(config=config)
|
||||
model.to(torch_device)
|
||||
model.eval()
|
||||
result = model(input_ids, attention_mask=input_mask)
|
||||
result = model(input_ids)
|
||||
self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size))
|
||||
|
||||
def create_and_check_model_as_decoder(
|
||||
self,
|
||||
config,
|
||||
input_ids,
|
||||
token_type_ids,
|
||||
input_mask,
|
||||
sequence_labels,
|
||||
token_labels,
|
||||
choice_labels,
|
||||
encoder_hidden_states,
|
||||
encoder_attention_mask,
|
||||
):
|
||||
config.add_cross_attention = True
|
||||
model = CohereModel(config)
|
||||
model.to(torch_device)
|
||||
model.eval()
|
||||
result = model(
|
||||
input_ids,
|
||||
attention_mask=input_mask,
|
||||
encoder_hidden_states=encoder_hidden_states,
|
||||
encoder_attention_mask=encoder_attention_mask,
|
||||
)
|
||||
result = model(
|
||||
input_ids,
|
||||
attention_mask=input_mask,
|
||||
encoder_hidden_states=encoder_hidden_states,
|
||||
)
|
||||
result = model(input_ids, attention_mask=input_mask)
|
||||
self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size))
|
||||
|
||||
def create_and_check_for_causal_lm(
|
||||
self,
|
||||
config,
|
||||
input_ids,
|
||||
token_type_ids,
|
||||
input_mask,
|
||||
sequence_labels,
|
||||
token_labels,
|
||||
choice_labels,
|
||||
encoder_hidden_states,
|
||||
encoder_attention_mask,
|
||||
):
|
||||
model = CohereForCausalLM(config=config)
|
||||
model.to(torch_device)
|
||||
model.eval()
|
||||
result = model(input_ids, attention_mask=input_mask, labels=token_labels)
|
||||
self.parent.assertEqual(result.logits.shape, (self.batch_size, self.seq_length, self.vocab_size))
|
||||
|
||||
def create_and_check_decoder_model_past_large_inputs(
|
||||
self,
|
||||
config,
|
||||
input_ids,
|
||||
token_type_ids,
|
||||
input_mask,
|
||||
sequence_labels,
|
||||
token_labels,
|
||||
choice_labels,
|
||||
encoder_hidden_states,
|
||||
encoder_attention_mask,
|
||||
):
|
||||
config.is_decoder = True
|
||||
config.add_cross_attention = True
|
||||
model = CohereForCausalLM(config=config)
|
||||
model.to(torch_device)
|
||||
model.eval()
|
||||
|
||||
# first forward pass
|
||||
outputs = model(
|
||||
input_ids,
|
||||
attention_mask=input_mask,
|
||||
encoder_hidden_states=encoder_hidden_states,
|
||||
encoder_attention_mask=encoder_attention_mask,
|
||||
use_cache=True,
|
||||
)
|
||||
past_key_values = outputs.past_key_values
|
||||
|
||||
# create hypothetical multiple next token and extent to next_input_ids
|
||||
next_tokens = ids_tensor((self.batch_size, 3), config.vocab_size)
|
||||
next_mask = ids_tensor((self.batch_size, 3), vocab_size=2)
|
||||
|
||||
# append to next input_ids and
|
||||
next_input_ids = torch.cat([input_ids, next_tokens], dim=-1)
|
||||
next_attention_mask = torch.cat([input_mask, next_mask], dim=-1)
|
||||
|
||||
output_from_no_past = model(
|
||||
next_input_ids,
|
||||
attention_mask=next_attention_mask,
|
||||
encoder_hidden_states=encoder_hidden_states,
|
||||
encoder_attention_mask=encoder_attention_mask,
|
||||
output_hidden_states=True,
|
||||
)["hidden_states"][0]
|
||||
output_from_past = model(
|
||||
next_tokens,
|
||||
attention_mask=next_attention_mask,
|
||||
encoder_hidden_states=encoder_hidden_states,
|
||||
encoder_attention_mask=encoder_attention_mask,
|
||||
past_key_values=past_key_values,
|
||||
output_hidden_states=True,
|
||||
)["hidden_states"][0]
|
||||
|
||||
# select random slice
|
||||
random_slice_idx = ids_tensor((1,), output_from_past.shape[-1]).item()
|
||||
output_from_no_past_slice = output_from_no_past[:, -3:, random_slice_idx].detach()
|
||||
output_from_past_slice = output_from_past[:, :, random_slice_idx].detach()
|
||||
|
||||
self.parent.assertTrue(output_from_past_slice.shape[1] == next_tokens.shape[1])
|
||||
|
||||
# test that outputs are equal for slice
|
||||
self.parent.assertTrue(torch.allclose(output_from_past_slice, output_from_no_past_slice, atol=1e-3))
|
||||
|
||||
def prepare_config_and_inputs_for_common(self):
|
||||
config_and_inputs = self.prepare_config_and_inputs()
|
||||
(
|
||||
config,
|
||||
input_ids,
|
||||
token_type_ids,
|
||||
input_mask,
|
||||
sequence_labels,
|
||||
token_labels,
|
||||
choice_labels,
|
||||
) = config_and_inputs
|
||||
inputs_dict = {"input_ids": input_ids, "attention_mask": input_mask}
|
||||
return config, inputs_dict
|
||||
|
||||
|
||||
@require_torch
|
||||
class CohereModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMixin, unittest.TestCase):
|
||||
all_model_classes = (CohereModel, CohereForCausalLM) if is_torch_available() else ()
|
||||
all_generative_model_classes = (CohereForCausalLM,) if is_torch_available() else ()
|
||||
pipeline_model_mapping = (
|
||||
{
|
||||
"feature-extraction": CohereModel,
|
||||
"text-generation": CohereForCausalLM,
|
||||
}
|
||||
if is_torch_available()
|
||||
else {}
|
||||
)
|
||||
test_headmasking = False
|
||||
test_pruning = False
|
||||
fx_compatible = True
|
||||
|
||||
# Need to use `0.8` instead of `0.9` for `test_cpu_offload`
|
||||
# This is because we are hitting edge cases with the causal_mask buffer
|
||||
model_split_percents = [0.5, 0.7, 0.8]
|
||||
|
||||
def setUp(self):
|
||||
self.model_tester = CohereModelTester(self)
|
||||
self.config_tester = ConfigTester(self, config_class=CohereConfig, hidden_size=37)
|
||||
|
||||
def test_config(self):
|
||||
self.config_tester.run_common_tests()
|
||||
|
||||
@unittest.skip("TODO @gante fix this for Cohere")
|
||||
@parameterized.expand([(1, False), (1, True), (4, False)])
|
||||
def test_new_cache_format(self, num_beams, do_sample):
|
||||
pass
|
||||
|
||||
def test_model(self):
|
||||
config_and_inputs = self.model_tester.prepare_config_and_inputs()
|
||||
self.model_tester.create_and_check_model(*config_and_inputs)
|
||||
|
||||
def test_model_various_embeddings(self):
|
||||
config_and_inputs = self.model_tester.prepare_config_and_inputs()
|
||||
for type in ["absolute", "relative_key", "relative_key_query"]:
|
||||
config_and_inputs[0].position_embedding_type = type
|
||||
self.model_tester.create_and_check_model(*config_and_inputs)
|
||||
|
||||
@require_bitsandbytes
|
||||
@require_torch_sdpa
|
||||
@require_torch_multi_gpu
|
||||
@slow
|
||||
def test_eager_matches_sdpa_generate(self):
|
||||
"""
|
||||
Overwritting the common test as the test is flaky on tiny models
|
||||
"""
|
||||
max_new_tokens = 30
|
||||
|
||||
model_id = "CohereForAI/c4ai-command-r-v01-4bit"
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
||||
|
||||
model_sdpa = CohereForCausalLM.from_pretrained(
|
||||
model_id, torch_dtype=torch.float16, low_cpu_mem_usage=True, device_map="auto"
|
||||
)
|
||||
self.assertTrue(model_sdpa.config._attn_implementation == "sdpa")
|
||||
|
||||
model_eager = CohereForCausalLM.from_pretrained(
|
||||
model_id, torch_dtype=torch.float16, attn_implementation="eager", device_map="auto"
|
||||
)
|
||||
|
||||
self.assertTrue(model_eager.config._attn_implementation == "eager")
|
||||
|
||||
for name, submodule in model_eager.named_modules():
|
||||
if "SdpaAttention" in submodule.__class__.__name__:
|
||||
raise ValueError("The eager model should not have SDPA attention layers")
|
||||
|
||||
has_sdpa = False
|
||||
for name, submodule in model_sdpa.named_modules():
|
||||
if "SdpaAttention" in submodule.__class__.__name__:
|
||||
has_sdpa = True
|
||||
break
|
||||
if not has_sdpa:
|
||||
raise ValueError("The SDPA model should have SDPA attention layers")
|
||||
|
||||
texts = [
|
||||
"hi here's a longer context, getting longer and",
|
||||
"Hello this is a very long sentence my friend, very long for real",
|
||||
"Today I am in Paris and",
|
||||
]
|
||||
|
||||
for padding_side in ["left", "right"]:
|
||||
tokenizer.padding_side = padding_side
|
||||
tokenizer.pad_token = tokenizer.eos_token
|
||||
|
||||
inputs = tokenizer(texts, return_tensors="pt", padding=True).to(torch_device)
|
||||
|
||||
res_eager = model_eager.generate(**inputs, max_new_tokens=max_new_tokens, do_sample=False)
|
||||
res_sdpa = model_sdpa.generate(**inputs, max_new_tokens=max_new_tokens, do_sample=False)
|
||||
|
||||
with self.subTest(f"{padding_side}"):
|
||||
torch.testing.assert_close(
|
||||
res_eager,
|
||||
res_sdpa,
|
||||
msg=f"\n{tokenizer.batch_decode(res_eager)} \nvs\n{tokenizer.batch_decode(res_sdpa)}",
|
||||
)
|
||||
|
||||
|
||||
@require_torch
|
||||
@slow
|
||||
class CohereIntegrationTest(unittest.TestCase):
|
||||
@require_torch_multi_gpu
|
||||
def test_batched_4bit(self):
|
||||
model_id = "CohereForAI/c4ai-command-r-v01-4bit"
|
||||
|
||||
EXPECTED_TEXT = [
|
||||
'Hello today I am going to show you how to make a simple and easy card using the new stamp set called "Hello" from the Occasions catalog. This set is so versatile and can be used for many occasions. I used the new In',
|
||||
"Hi there, here we are again with another great collection of free fonts. This time we have gathered 10 free fonts that you can download and use in your designs. These fonts are free for personal and commercial use. So",
|
||||
]
|
||||
|
||||
model = CohereForCausalLM.from_pretrained(model_id)
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
||||
|
||||
tokenizer.pad_token = tokenizer.eos_token
|
||||
|
||||
text = ["Hello today I am going to show you how to", "Hi there, here we are"]
|
||||
inputs = tokenizer(text, return_tensors="pt", padding=True).to(torch_device)
|
||||
|
||||
output = model.generate(**inputs, max_new_tokens=40, do_sample=False)
|
||||
self.assertEqual(tokenizer.batch_decode(output, skip_special_tokens=True), EXPECTED_TEXT)
|
||||
|
||||
def test_batched_small_model_logits(self):
|
||||
# Since the model is very large, we created a random cohere model so that we can do a simple
|
||||
# logits check on it.
|
||||
model_id = "hf-internal-testing/cohere-random"
|
||||
|
||||
EXPECTED_LOGITS = torch.Tensor(
|
||||
[
|
||||
[[0.0000, 0.1866, -0.1997], [0.0000, -0.0736, 0.1785], [0.0000, -0.1965, -0.0569]],
|
||||
[[0.0000, -0.0302, 0.1488], [0.0000, -0.0402, 0.1351], [0.0000, -0.0341, 0.1116]],
|
||||
]
|
||||
).to(torch_device)
|
||||
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
||||
model = CohereForCausalLM.from_pretrained(model_id, low_cpu_mem_usage=True, torch_dtype=torch.float16).to(
|
||||
torch_device
|
||||
)
|
||||
|
||||
tokenizer.pad_token = tokenizer.eos_token
|
||||
|
||||
text = ["Hello today I am going to show you how to", "Hi there, here we are"]
|
||||
inputs = tokenizer(text, return_tensors="pt", padding=True).to(torch_device)
|
||||
|
||||
with torch.no_grad():
|
||||
output = model(**inputs)
|
||||
|
||||
logits = output.logits
|
||||
self.assertTrue(torch.allclose(EXPECTED_LOGITS, logits[:, :3, :3], rtol=1e-3, atol=1e-3))
|
|
@ -0,0 +1,297 @@
|
|||
# coding=utf-8
|
||||
# Copyright 2022 The HuggingFace Team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
import unittest
|
||||
|
||||
from transformers import CohereTokenizerFast
|
||||
from transformers.testing_utils import require_jinja, require_tokenizers
|
||||
|
||||
from ...test_tokenization_common import TokenizerTesterMixin
|
||||
|
||||
|
||||
@require_tokenizers
|
||||
class CohereTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
|
||||
slow_tokenizer_class = None
|
||||
rust_tokenizer_class = CohereTokenizerFast
|
||||
tokenizer_class = CohereTokenizerFast
|
||||
test_rust_tokenizer = True
|
||||
test_slow_tokenizer = False
|
||||
from_pretrained_vocab_key = "tokenizer_file"
|
||||
from_pretrained_id = "CohereForAI/c4ai-command-r-v01"
|
||||
special_tokens_map = {
|
||||
"bos_token": "<BOS_TOKEN>",
|
||||
"eos_token": "<|END_OF_TURN_TOKEN|>",
|
||||
"unk_token": "<UNK>",
|
||||
"pad_token": "<PAD>",
|
||||
}
|
||||
|
||||
def setUp(self):
|
||||
super().setUp()
|
||||
tokenizer = CohereTokenizerFast.from_pretrained("CohereForAI/c4ai-command-r-v01")
|
||||
tokenizer.save_pretrained(self.tmpdirname)
|
||||
|
||||
def get_rust_tokenizer(self, **kwargs):
|
||||
kwargs.update(self.special_tokens_map)
|
||||
return CohereTokenizerFast.from_pretrained(self.tmpdirname, **kwargs)
|
||||
|
||||
@unittest.skip("This needs a slow tokenizer. Cohere does not have one!")
|
||||
def test_encode_decode_with_spaces(self):
|
||||
return
|
||||
|
||||
def test_encodings_from_sample_data(self):
|
||||
"""
|
||||
Assert that the created tokens are the same than the hard-coded ones
|
||||
"""
|
||||
tokenizer = self.get_rust_tokenizer()
|
||||
|
||||
INPUT_SENTENCES = ["The quick brown fox<|END_OF_TURN_TOKEN|>", "jumps over the lazy dog<|END_OF_TURN_TOKEN|>"]
|
||||
TARGET_TOKENS = [[5, 2162, 6629, 19883, 73388, 255001], [5, 81, 25092, 2515, 1690, 46189, 9507, 255001]]
|
||||
|
||||
computed_tokens = tokenizer.batch_encode_plus(INPUT_SENTENCES)["input_ids"]
|
||||
self.assertListEqual(TARGET_TOKENS, computed_tokens)
|
||||
|
||||
INPUT_SENTENCES_W_BOS = [
|
||||
"<BOS_TOKEN>The quick brown fox<|END_OF_TURN_TOKEN|>",
|
||||
"<BOS_TOKEN>jumps over the lazy dog<|END_OF_TURN_TOKEN|>",
|
||||
]
|
||||
decoded_tokens = tokenizer.batch_decode(computed_tokens)
|
||||
self.assertListEqual(decoded_tokens, INPUT_SENTENCES_W_BOS)
|
||||
|
||||
def test_padding(self, max_length=10):
|
||||
for tokenizer, pretrained_name, kwargs in self.tokenizers_list:
|
||||
with self.subTest(f"{tokenizer.__class__.__name__} ({pretrained_name})"):
|
||||
tokenizer_r = self.rust_tokenizer_class.from_pretrained(pretrained_name, **kwargs)
|
||||
# tokenizer_r.pad_token = None # Hotfixing padding = None
|
||||
# Simple input
|
||||
s = "This is a simple input"
|
||||
s2 = ["This is a simple input 1", "This is a simple input 2"]
|
||||
p = ("This is a simple input", "This is a pair")
|
||||
p2 = [
|
||||
("This is a simple input 1", "This is a simple input 2"),
|
||||
("This is a simple pair 1", "This is a simple pair 2"),
|
||||
]
|
||||
|
||||
# Simple input tests
|
||||
try:
|
||||
tokenizer_r.encode(s, max_length=max_length)
|
||||
tokenizer_r.encode_plus(s, max_length=max_length)
|
||||
|
||||
tokenizer_r.batch_encode_plus(s2, max_length=max_length)
|
||||
tokenizer_r.encode(p, max_length=max_length)
|
||||
tokenizer_r.batch_encode_plus(p2, max_length=max_length)
|
||||
except ValueError:
|
||||
self.fail("Cohere Tokenizer should be able to deal with padding")
|
||||
|
||||
tokenizer_r.pad_token = None # Hotfixing padding = None
|
||||
self.assertRaises(ValueError, tokenizer_r.encode, s, max_length=max_length, padding="max_length")
|
||||
|
||||
# Simple input
|
||||
self.assertRaises(ValueError, tokenizer_r.encode_plus, s, max_length=max_length, padding="max_length")
|
||||
|
||||
# Simple input
|
||||
self.assertRaises(
|
||||
ValueError,
|
||||
tokenizer_r.batch_encode_plus,
|
||||
s2,
|
||||
max_length=max_length,
|
||||
padding="max_length",
|
||||
)
|
||||
|
||||
# Pair input
|
||||
self.assertRaises(ValueError, tokenizer_r.encode, p, max_length=max_length, padding="max_length")
|
||||
|
||||
# Pair input
|
||||
self.assertRaises(ValueError, tokenizer_r.encode_plus, p, max_length=max_length, padding="max_length")
|
||||
|
||||
# Pair input
|
||||
self.assertRaises(
|
||||
ValueError,
|
||||
tokenizer_r.batch_encode_plus,
|
||||
p2,
|
||||
max_length=max_length,
|
||||
padding="max_length",
|
||||
)
|
||||
|
||||
def test_pretrained_model_lists(self):
|
||||
# No `max_model_input_sizes` for Cohere model
|
||||
self.assertGreaterEqual(len(self.tokenizer_class.pretrained_vocab_files_map), 1)
|
||||
self.assertGreaterEqual(len(list(self.tokenizer_class.pretrained_vocab_files_map.values())[0]), 1)
|
||||
|
||||
@require_jinja
|
||||
def test_tokenization_for_chat(self):
|
||||
tokenizer = self.get_rust_tokenizer()
|
||||
test_chats = [
|
||||
[{"role": "system", "content": "You are a helpful chatbot."}, {"role": "user", "content": "Hello!"}],
|
||||
[
|
||||
{"role": "system", "content": "You are a helpful chatbot."},
|
||||
{"role": "user", "content": "Hello!"},
|
||||
{"role": "assistant", "content": "Nice to meet you."},
|
||||
],
|
||||
]
|
||||
tokenized_chats = [tokenizer.apply_chat_template(test_chat) for test_chat in test_chats]
|
||||
expected_tokens = [
|
||||
[5, 255000, 255008, 5659, 1955, 1671, 19264, 171597, 21, 255001, 255000, 255006, 28339, 8, 255001],
|
||||
[
|
||||
5,
|
||||
255000,
|
||||
255008,
|
||||
5659,
|
||||
1955,
|
||||
1671,
|
||||
19264,
|
||||
171597,
|
||||
21,
|
||||
255001,
|
||||
255000,
|
||||
255006,
|
||||
28339,
|
||||
8,
|
||||
255001,
|
||||
255000,
|
||||
255007,
|
||||
97190,
|
||||
1726,
|
||||
5694,
|
||||
1933,
|
||||
21,
|
||||
255001,
|
||||
],
|
||||
]
|
||||
for tokenized_chat, expected_tokens in zip(tokenized_chats, expected_tokens):
|
||||
self.assertListEqual(tokenized_chat, expected_tokens)
|
||||
|
||||
@require_jinja
|
||||
def test_tokenization_for_tool_use(self):
|
||||
tokenizer = self.get_rust_tokenizer()
|
||||
|
||||
conversation = [{"role": "user", "content": "Whats the biggest penguin in the world?"}]
|
||||
|
||||
tools = [
|
||||
{
|
||||
"name": "internet_search",
|
||||
"description": "Returns a list of relevant document snippets for a textual query retrieved from the internet",
|
||||
"parameter_definitions": {
|
||||
"query": {"description": "Query to search the internet with", "type": "str", "required": True}
|
||||
},
|
||||
},
|
||||
{
|
||||
"name": "directly_answer",
|
||||
"description": "Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history",
|
||||
"parameter_definitions": {},
|
||||
},
|
||||
]
|
||||
|
||||
tool_use_prompt = tokenizer.apply_tool_use_template(
|
||||
conversation,
|
||||
tools=tools,
|
||||
tokenize=False,
|
||||
add_generation_prompt=True,
|
||||
)
|
||||
|
||||
expected_prompt = '''<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
|
||||
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
|
||||
|
||||
# System Preamble
|
||||
## Basic Rules
|
||||
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
|
||||
|
||||
# User Preamble
|
||||
## Task and Context
|
||||
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
|
||||
|
||||
## Style Guide
|
||||
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.
|
||||
|
||||
## Available Tools
|
||||
Here is a list of tools that you have available to you:
|
||||
|
||||
```python
|
||||
def internet_search(query: str) -> List[Dict]:
|
||||
"""Returns a list of relevant document snippets for a textual query retrieved from the internet
|
||||
|
||||
Args:
|
||||
query (str): Query to search the internet with
|
||||
"""
|
||||
pass
|
||||
```
|
||||
|
||||
```python
|
||||
def directly_answer() -> List[Dict]:
|
||||
"""Calls a standard (un-augmented) AI chatbot to generate a response given the conversation history
|
||||
"""
|
||||
pass
|
||||
```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Write 'Action:' followed by a json-formatted list of actions that you want to perform in order to produce a good response to the user's last input. You can use any of the supplied tools any number of times, but you should aim to execute the minimum number of necessary actions for the input. You should use the `directly-answer` tool if calling the other tools is unnecessary. The list of actions you want to call should be formatted as a list of json objects, for example:
|
||||
```json
|
||||
[
|
||||
{
|
||||
"tool_name": title of the tool in the specification,
|
||||
"parameters": a dict of parameters to input into the tool as they are defined in the specs, or {} if it takes no parameters
|
||||
}
|
||||
]```<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>'''
|
||||
|
||||
self.assertEqual(tool_use_prompt, expected_prompt)
|
||||
|
||||
@require_jinja
|
||||
def test_tokenization_for_grounded_generation(self):
|
||||
tokenizer = self.get_rust_tokenizer()
|
||||
conversation = [{"role": "user", "content": "Whats the biggest penguin in the world?"}]
|
||||
|
||||
documents = [
|
||||
{"title": "Tall penguins", "text": "Emperor penguins are the tallest growing up to 122 cm in height."},
|
||||
{"title": "Penguin habitats", "text": "Emperor penguins only live in Antarctica."},
|
||||
]
|
||||
|
||||
grounded_generation_prompt = tokenizer.apply_grounded_generation_template(
|
||||
conversation,
|
||||
documents=documents,
|
||||
citation_mode="accurate", # or "fast"
|
||||
tokenize=False,
|
||||
add_generation_prompt=True,
|
||||
)
|
||||
|
||||
expected_prompt = """<BOS_TOKEN><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|># Safety Preamble
|
||||
The instructions in this section override those in the task description and style guide sections. Don't answer questions that are harmful or immoral.
|
||||
|
||||
# System Preamble
|
||||
## Basic Rules
|
||||
You are a powerful conversational AI trained by Cohere to help people. You are augmented by a number of tools, and your job is to use and consume the output of these tools to best help the user. You will see a conversation history between yourself and a user, ending with an utterance from the user. You will then see a specific instruction instructing you what kind of response to generate. When you answer the user's requests, you cite your sources in your answers, according to those instructions.
|
||||
|
||||
# User Preamble
|
||||
## Task and Context
|
||||
You help people answer their questions and other requests interactively. You will be asked a very wide array of requests on all kinds of topics. You will be equipped with a wide range of search engines or similar tools to help you, which you use to research your answer. You should focus on serving the user's needs as best you can, which will be wide-ranging.
|
||||
|
||||
## Style Guide
|
||||
Unless the user asks for a different style of answer, you should answer in full sentences, using proper grammar and spelling.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Whats the biggest penguin in the world?<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|><results>
|
||||
Document: 0
|
||||
title: Tall penguins
|
||||
text: Emperor penguins are the tallest growing up to 122 cm in height.
|
||||
|
||||
Document: 1
|
||||
title: Penguin habitats
|
||||
text: Emperor penguins only live in Antarctica.
|
||||
</results><|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>Carefully perform the following instructions, in order, starting each with a new line.
|
||||
Firstly, Decide which of the retrieved documents are relevant to the user's last input by writing 'Relevant Documents:' followed by comma-separated list of document numbers. If none are relevant, you should instead write 'None'.
|
||||
Secondly, Decide which of the retrieved documents contain facts that should be cited in a good answer to the user's last input by writing 'Cited Documents:' followed a comma-separated list of document numbers. If you dont want to cite any of them, you should instead write 'None'.
|
||||
Thirdly, Write 'Answer:' followed by a response to the user's last input in high quality natural english. Use the retrieved documents to help you. Do not insert any citations or grounding markup.
|
||||
Finally, Write 'Grounded answer:' followed by a response to the user's last input in high quality natural english. Use the symbols <co: doc> and </co: doc> to indicate when a fact comes from a document in the search result, e.g <co: 0>my fact</co: 0> for a fact from document 0.<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>"""
|
||||
|
||||
self.assertEqual(grounded_generation_prompt, expected_prompt)
|
||||
|
||||
def test_add_prefix_space_fast(self):
|
||||
tokenizer_w_prefix = self.get_rust_tokenizer(add_prefix_space=True)
|
||||
tokenizer_wo_prefix = self.get_rust_tokenizer(add_prefix_space=False)
|
||||
tokens_w_prefix = tokenizer_w_prefix.tokenize("Hey")
|
||||
tokens_wo_prefix = tokenizer_wo_prefix.tokenize("Hey")
|
||||
self.assertNotEqual(tokens_w_prefix, tokens_wo_prefix)
|
Loading…
Reference in New Issue