vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

refactor: Add proper names for remote model references

Browse source
This commit is contained in:
teleprint-me 2024-05-20 01:36:44 -04:00
parent 6fc4492b3f
commit a1951e27dc
No known key found for this signature in database
GPG key ID: B0D11345E65C4D48
1 changed files with 22 additions and 22 deletions
Download patch file Download diff file Expand all files Collapse all files

44
gguf-py/gguf/huggingface_hub.py
View file

@ -29,28 +29,28 @@ class LLaMaModelType(IntEnum):
# - Use architecture types because they are explicitly defined
# - Possible tokenizer model types are: SentencePiece, WordPiece, or BytePair
MODELS = (
{"arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.SPM, "repo": "meta-llama/Llama-2-7b-hf", },
{"arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.BPE, "repo": "meta-llama/Meta-Llama-3-8B", },
{"arch": MODEL_ARCH.PHI3, "vocab_type": LLaMaVocabType.SPM, "repo": "microsoft/Phi-3-mini-4k-instruct", },
{"arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.BPE, "repo": "deepseek-ai/deepseek-llm-7b-base", },
{"arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.BPE, "repo": "deepseek-ai/deepseek-coder-6.7b-base", },
{"arch": MODEL_ARCH.FALCON, "vocab_type": LLaMaVocabType.BPE, "repo": "tiiuae/falcon-7b", },
{"arch": MODEL_ARCH.BERT, "vocab_type": LLaMaVocabType.WPM, "repo": "BAAI/bge-small-en-v1.5", },
{"arch": MODEL_ARCH.MPT, "vocab_type": LLaMaVocabType.BPE, "repo": "mosaicml/mpt-7b", },
{"arch": MODEL_ARCH.STARCODER2, "vocab_type": LLaMaVocabType.BPE, "repo": "bigcode/starcoder2-3b", },
{"arch": MODEL_ARCH.GPT2, "vocab_type": LLaMaVocabType.BPE, "repo": "openai-community/gpt2", },
{"arch": MODEL_ARCH.REFACT, "vocab_type": LLaMaVocabType.BPE, "repo": "smallcloudai/Refact-1_6-base", },
{"arch": MODEL_ARCH.COMMAND_R, "vocab_type": LLaMaVocabType.BPE, "repo": "CohereForAI/c4ai-command-r-v01", },
{"arch": MODEL_ARCH.QWEN2, "vocab_type": LLaMaVocabType.BPE, "repo": "Qwen/Qwen1.5-7B", },
{"arch": MODEL_ARCH.OLMO, "vocab_type": LLaMaVocabType.BPE, "repo": "allenai/OLMo-1.7-7B-hf", },
{"arch": MODEL_ARCH.DBRX, "vocab_type": LLaMaVocabType.BPE, "repo": "databricks/dbrx-base", },
{"arch": MODEL_ARCH.JINA_BERT_V2, "vocab_type": LLaMaVocabType.WPM, "repo": "jinaai/jina-embeddings-v2-base-en", },
{"arch": MODEL_ARCH.JINA_BERT_V2, "vocab_type": LLaMaVocabType.BPE, "repo": "jinaai/jina-embeddings-v2-base-es", },
{"arch": MODEL_ARCH.JINA_BERT_V2, "vocab_type": LLaMaVocabType.BPE, "repo": "jinaai/jina-embeddings-v2-base-de", },
{"arch": MODEL_ARCH.PHI2, "vocab_type": LLaMaVocabType.BPE, "repo": "microsoft/phi-1", },
{"arch": MODEL_ARCH.STABLELM, "vocab_type": LLaMaVocabType.BPE, "repo": "stabilityai/stablelm-2-zephyr-1_6b", },
{"arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.SPM, "repo": "mistralai/Mistral-7B-Instruct-v0.2", },
{"arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.SPM, "repo": "mistralai/Mixtral-8x7B-Instruct-v0.1", },
{"model_arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.SPM, "repo": "meta-llama/Llama-2-7b-hf", },
{"model_arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.BPE, "repo": "meta-llama/Meta-Llama-3-8B", },
{"model_arch": MODEL_ARCH.PHI3, "vocab_type": LLaMaVocabType.SPM, "repo": "microsoft/Phi-3-mini-4k-instruct", },
{"model_arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.BPE, "repo": "deepseek-ai/deepseek-llm-7b-base", },
{"model_arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.BPE, "repo": "deepseek-ai/deepseek-coder-6.7b-base", },
{"model_arch": MODEL_ARCH.FALCON, "vocab_type": LLaMaVocabType.BPE, "repo": "tiiuae/falcon-7b", },
{"model_arch": MODEL_ARCH.BERT, "vocab_type": LLaMaVocabType.WPM, "repo": "BAAI/bge-small-en-v1.5", },
{"model_arch": MODEL_ARCH.MPT, "vocab_type": LLaMaVocabType.BPE, "repo": "mosaicml/mpt-7b", },
{"model_arch": MODEL_ARCH.STARCODER2, "vocab_type": LLaMaVocabType.BPE, "repo": "bigcode/starcoder2-3b", },
{"model_arch": MODEL_ARCH.GPT2, "vocab_type": LLaMaVocabType.BPE, "repo": "openai-community/gpt2", },
{"model_arch": MODEL_ARCH.REFACT, "vocab_type": LLaMaVocabType.BPE, "repo": "smallcloudai/Refact-1_6-base", },
{"model_arch": MODEL_ARCH.COMMAND_R, "vocab_type": LLaMaVocabType.BPE, "repo": "CohereForAI/c4ai-command-r-v01", },
{"model_arch": MODEL_ARCH.QWEN2, "vocab_type": LLaMaVocabType.BPE, "repo": "Qwen/Qwen1.5-7B", },
{"model_arch": MODEL_ARCH.OLMO, "vocab_type": LLaMaVocabType.BPE, "repo": "allenai/OLMo-1.7-7B-hf", },
{"model_arch": MODEL_ARCH.DBRX, "vocab_type": LLaMaVocabType.BPE, "repo": "databricks/dbrx-base", },
{"model_arch": MODEL_ARCH.JINA_BERT_V2, "vocab_type": LLaMaVocabType.WPM, "repo": "jinaai/jina-embeddings-v2-base-en", },
{"model_arch": MODEL_ARCH.JINA_BERT_V2, "vocab_type": LLaMaVocabType.BPE, "repo": "jinaai/jina-embeddings-v2-base-es", },
{"model_arch": MODEL_ARCH.JINA_BERT_V2, "vocab_type": LLaMaVocabType.BPE, "repo": "jinaai/jina-embeddings-v2-base-de", },
{"model_arch": MODEL_ARCH.PHI2, "vocab_type": LLaMaVocabType.BPE, "repo": "microsoft/phi-1", },
{"model_arch": MODEL_ARCH.STABLELM, "vocab_type": LLaMaVocabType.BPE, "repo": "stabilityai/stablelm-2-zephyr-1_6b", },
{"model_arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.SPM, "repo": "mistralai/Mistral-7B-Instruct-v0.2", },
{"model_arch": MODEL_ARCH.LLAMA, "vocab_type": LLaMaVocabType.SPM, "repo": "mistralai/Mixtral-8x7B-Instruct-v0.1", },
)