commit 9fb1a0f077b86f8656884d5bf0391d7127e0755d Author: ModelHub XC Date: Sat May 2 22:36:16 2026 +0800 初始化项目，由ModelHub XC社区提供模型 Model: double7/vicuna-68m Source: Original Platform diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..a6344aa --- /dev/null +++ b/.gitattributes @@ -0,0 +1,35 @@ +*.7z filter=lfs diff=lfs merge=lfs -text +*.arrow filter=lfs diff=lfs merge=lfs -text +*.bin filter=lfs diff=lfs merge=lfs -text +*.bz2 filter=lfs diff=lfs merge=lfs -text +*.ckpt filter=lfs diff=lfs merge=lfs -text +*.ftz filter=lfs diff=lfs merge=lfs -text +*.gz filter=lfs diff=lfs merge=lfs -text +*.h5 filter=lfs diff=lfs merge=lfs -text +*.joblib filter=lfs diff=lfs merge=lfs -text +*.lfs.* filter=lfs diff=lfs merge=lfs -text +*.mlmodel filter=lfs diff=lfs merge=lfs -text +*.model filter=lfs diff=lfs merge=lfs -text +*.msgpack filter=lfs diff=lfs merge=lfs -text +*.npy filter=lfs diff=lfs merge=lfs -text +*.npz filter=lfs diff=lfs merge=lfs -text +*.onnx filter=lfs diff=lfs merge=lfs -text +*.ot filter=lfs diff=lfs merge=lfs -text +*.parquet filter=lfs diff=lfs merge=lfs -text +*.pb filter=lfs diff=lfs merge=lfs -text +*.pickle filter=lfs diff=lfs merge=lfs -text +*.pkl filter=lfs diff=lfs merge=lfs -text +*.pt filter=lfs diff=lfs merge=lfs -text +*.pth filter=lfs diff=lfs merge=lfs -text +*.rar filter=lfs diff=lfs merge=lfs -text +*.safetensors filter=lfs diff=lfs merge=lfs -text +saved_model/**/* filter=lfs diff=lfs merge=lfs -text +*.tar.* filter=lfs diff=lfs merge=lfs -text +*.tar filter=lfs diff=lfs merge=lfs -text +*.tflite filter=lfs diff=lfs merge=lfs -text +*.tgz filter=lfs diff=lfs merge=lfs -text +*.wasm filter=lfs diff=lfs merge=lfs -text +*.xz filter=lfs diff=lfs merge=lfs -text +*.zip filter=lfs diff=lfs merge=lfs -text +*.zst filter=lfs diff=lfs merge=lfs -text +*tfevents* filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..376b5c9 --- /dev/null +++ b/README.md @@ -0,0 +1,22 @@ +--- +license: apache-2.0 +datasets: +- anon8231489123/ShareGPT_Vicuna_unfiltered +language: +- en +pipeline_tag: text-generation +--- +## Model description +This is a Vicuna-like model with only 68M parameters, which is fine-tuned from [LLaMA-68m](https://huggingface.co/JackFram/llama-68m) on [ShareGPT](https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered) data. + +The training setup follows the [Vicuna suite](https://github.com/lm-sys/FastChat). + +The model is mainly developed as a base Small Speculative Model in the [MCSD paper](https://arxiv.org/pdf/2401.06706.pdf). As a comparison, it can be better aligned to the Vicuna models than LLaMA-68m with little loss of alignment to the LLaMA models. + + +| Draft Model | Target Model | Alignment | +| -------------- | ------------- | --------- | +| LLaMA-68/160M | LLaMA-13/33B | 😃 | +| LLaMA-68/160M | Vicuna-13/33B | 😟 | +| Vicuna-68/160M | LLaMA-13/33B | 😃 | +| Vicuna-68/160M | Vicuna-13/33B | 😃 | \ No newline at end of file diff --git a/config.json b/config.json new file mode 100644 index 0000000..2441b1f --- /dev/null +++ b/config.json @@ -0,0 +1,29 @@ +{ + "_name_or_path": "/home/nfs03/yangs/LLM/llama-68m", + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 0, + "eos_token_id": 2, + "hidden_act": "silu", + "hidden_size": 768, + "initializer_range": 0.02, + "intermediate_size": 3072, + "max_position_embeddings": 2048, + "model_type": "llama", + "num_attention_heads": 12, + "num_hidden_layers": 2, + "num_key_value_heads": 12, + "pad_token_id": 1, + "pretraining_tp": 1, + "rms_norm_eps": 1e-06, + "rope_scaling": null, + "rope_theta": 10000.0, + "tie_word_embeddings": false, + "torch_dtype": "float32", + "transformers_version": "4.37.0.dev0", + "use_cache": true, + "vocab_size": 32000 +} diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000..d6af2fe --- /dev/null +++ b/generation_config.json @@ -0,0 +1,7 @@ +{ + "_from_model_config": true, + "bos_token_id": 0, + "eos_token_id": 2, + "pad_token_id": 1, + "transformers_version": "4.37.0.dev0" +} diff --git a/model.safetensors b/model.safetensors new file mode 100644 index 0000000..61c6883 --- /dev/null +++ b/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:049aa4dc91bd8bb0e3009b8408b59512d8e8eb6aced19346570cc82fa657113d +size 272123144 diff --git a/special_tokens_map.json b/special_tokens_map.json new file mode 100644 index 0000000..f928b24 --- /dev/null +++ b/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + } +} diff --git a/tokenizer.model b/tokenizer.model new file mode 100644 index 0000000..6c00c74 --- /dev/null +++ b/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000..58c0f19 --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,43 @@ +{ + "add_bos_token": true, + "add_eos_token": false, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "legacy": true, + "model_max_length": 2048, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false, + "use_fast": true +} diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000..4ca8009 --- /dev/null +++ b/trainer_state.json @@ -0,0 +1,546 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 500, + "global_step": 4323, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.03, + "learning_rate": 7.692307692307694e-06, + "loss": 3.6444, + "step": 50 + }, + { + "epoch": 0.07, + "learning_rate": 1.5384615384615387e-05, + "loss": 3.0454, + "step": 100 + }, + { + "epoch": 0.1, + "learning_rate": 1.9998877278887434e-05, + "loss": 2.7656, + "step": 150 + }, + { + "epoch": 0.14, + "learning_rate": 1.9986249561409415e-05, + "loss": 2.5791, + "step": 200 + }, + { + "epoch": 0.17, + "learning_rate": 1.9959608504164855e-05, + "loss": 2.4997, + "step": 250 + }, + { + "epoch": 0.21, + "learning_rate": 1.9918991491577664e-05, + "loss": 2.4208, + "step": 300 + }, + { + "epoch": 0.24, + "learning_rate": 1.9864455520020126e-05, + "loss": 2.3673, + "step": 350 + }, + { + "epoch": 0.28, + "learning_rate": 1.9796077117831955e-05, + "loss": 2.3077, + "step": 400 + }, + { + "epoch": 0.31, + "learning_rate": 1.971395223793088e-05, + "loss": 2.2695, + "step": 450 + }, + { + "epoch": 0.35, + "learning_rate": 1.9618196123165404e-05, + "loss": 2.2326, + "step": 500 + }, + { + "epoch": 0.38, + "learning_rate": 1.9508943144598726e-05, + "loss": 2.2122, + "step": 550 + }, + { + "epoch": 0.42, + "learning_rate": 1.9386346612950727e-05, + "loss": 2.1661, + "step": 600 + }, + { + "epoch": 0.45, + "learning_rate": 1.925057856346263e-05, + "loss": 2.1537, + "step": 650 + }, + { + "epoch": 0.49, + "learning_rate": 1.9101829514486206e-05, + "loss": 2.1359, + "step": 700 + }, + { + "epoch": 0.52, + "learning_rate": 1.8940308200136353e-05, + "loss": 2.0954, + "step": 750 + }, + { + "epoch": 0.56, + "learning_rate": 1.8766241277382096e-05, + "loss": 2.0736, + "step": 800 + }, + { + "epoch": 0.59, + "learning_rate": 1.8579873007987135e-05, + "loss": 2.0633, + "step": 850 + }, + { + "epoch": 0.62, + "learning_rate": 1.838146491574624e-05, + "loss": 2.0746, + "step": 900 + }, + { + "epoch": 0.66, + "learning_rate": 1.8171295419498417e-05, + "loss": 2.0407, + "step": 950 + }, + { + "epoch": 0.69, + "learning_rate": 1.794965944243194e-05, + "loss": 2.0361, + "step": 1000 + }, + { + "epoch": 0.73, + "learning_rate": 1.7716867998229368e-05, + "loss": 2.0215, + "step": 1050 + }, + { + "epoch": 0.76, + "learning_rate": 1.747324775463342e-05, + "loss": 1.9921, + "step": 1100 + }, + { + "epoch": 0.8, + "learning_rate": 1.721914057504601e-05, + "loss": 2.0012, + "step": 1150 + }, + { + "epoch": 0.83, + "learning_rate": 1.695490303880384e-05, + "loss": 1.9912, + "step": 1200 + }, + { + "epoch": 0.87, + "learning_rate": 1.6680905940803596e-05, + "loss": 1.9723, + "step": 1250 + }, + { + "epoch": 0.9, + "learning_rate": 1.6397533771179006e-05, + "loss": 1.9712, + "step": 1300 + }, + { + "epoch": 0.94, + "learning_rate": 1.6105184175759864e-05, + "loss": 1.9479, + "step": 1350 + }, + { + "epoch": 0.97, + "learning_rate": 1.5804267398070123e-05, + "loss": 1.9513, + "step": 1400 + }, + { + "epoch": 1.01, + "learning_rate": 1.549520570364815e-05, + "loss": 1.9449, + "step": 1450 + }, + { + "epoch": 1.04, + "learning_rate": 1.5178432787496863e-05, + "loss": 1.9101, + "step": 1500 + }, + { + "epoch": 1.08, + "learning_rate": 1.4854393165495381e-05, + "loss": 1.9104, + "step": 1550 + }, + { + "epoch": 1.11, + "learning_rate": 1.4523541550626093e-05, + "loss": 1.8977, + "step": 1600 + }, + { + "epoch": 1.15, + "learning_rate": 1.4186342214892513e-05, + "loss": 1.9062, + "step": 1650 + }, + { + "epoch": 1.18, + "learning_rate": 1.3843268337823312e-05, + "loss": 1.9024, + "step": 1700 + }, + { + "epoch": 1.21, + "learning_rate": 1.3494801342476752e-05, + "loss": 1.9043, + "step": 1750 + }, + { + "epoch": 1.25, + "learning_rate": 1.3141430219877257e-05, + "loss": 1.8892, + "step": 1800 + }, + { + "epoch": 1.28, + "learning_rate": 1.2783650842832161e-05, + "loss": 1.884, + "step": 1850 + }, + { + "epoch": 1.32, + "learning_rate": 1.2421965270091454e-05, + "loss": 1.8725, + "step": 1900 + }, + { + "epoch": 1.35, + "learning_rate": 1.205688104182709e-05, + "loss": 1.8847, + "step": 1950 + }, + { + "epoch": 1.39, + "learning_rate": 1.168891046742033e-05, + "loss": 1.8765, + "step": 2000 + }, + { + "epoch": 1.42, + "learning_rate": 1.1318569906556733e-05, + "loss": 1.8787, + "step": 2050 + }, + { + "epoch": 1.46, + "learning_rate": 1.0946379044637374e-05, + "loss": 1.8805, + "step": 2100 + }, + { + "epoch": 1.49, + "learning_rate": 1.0572860163523335e-05, + "loss": 1.8752, + "step": 2150 + }, + { + "epoch": 1.53, + "learning_rate": 1.0198537408636583e-05, + "loss": 1.8843, + "step": 2200 + }, + { + "epoch": 1.56, + "learning_rate": 9.823936053445839e-06, + "loss": 1.8582, + "step": 2250 + }, + { + "epoch": 1.6, + "learning_rate": 9.449581762369454e-06, + "loss": 1.8684, + "step": 2300 + }, + { + "epoch": 1.63, + "learning_rate": 9.075999853129733e-06, + "loss": 1.8533, + "step": 2350 + }, + { + "epoch": 1.67, + "learning_rate": 8.703714559593719e-06, + "loss": 1.8464, + "step": 2400 + }, + { + "epoch": 1.7, + "learning_rate": 8.333248296134922e-06, + "loss": 1.8423, + "step": 2450 + }, + { + "epoch": 1.73, + "learning_rate": 7.965120924548302e-06, + "loss": 1.8669, + "step": 2500 + }, + { + "epoch": 1.77, + "learning_rate": 7.599849024547167e-06, + "loss": 1.8493, + "step": 2550 + }, + { + "epoch": 1.8, + "learning_rate": 7.237945168865693e-06, + "loss": 1.8523, + "step": 2600 + }, + { + "epoch": 1.84, + "learning_rate": 6.879917203984306e-06, + "loss": 1.8661, + "step": 2650 + }, + { + "epoch": 1.87, + "learning_rate": 6.526267537487263e-06, + "loss": 1.8445, + "step": 2700 + }, + { + "epoch": 1.91, + "learning_rate": 6.177492433052374e-06, + "loss": 1.8412, + "step": 2750 + }, + { + "epoch": 1.94, + "learning_rate": 5.834081314062314e-06, + "loss": 1.8391, + "step": 2800 + }, + { + "epoch": 1.98, + "learning_rate": 5.496516076814623e-06, + "loss": 1.8479, + "step": 2850 + }, + { + "epoch": 2.01, + "learning_rate": 5.165270414294229e-06, + "loss": 1.8553, + "step": 2900 + }, + { + "epoch": 2.05, + "learning_rate": 4.8408091514573454e-06, + "loss": 1.8322, + "step": 2950 + }, + { + "epoch": 2.08, + "learning_rate": 4.523587592959557e-06, + "loss": 1.8343, + "step": 3000 + }, + { + "epoch": 2.12, + "learning_rate": 4.214050884243429e-06, + "loss": 1.8277, + "step": 3050 + }, + { + "epoch": 2.15, + "learning_rate": 3.912633386882113e-06, + "loss": 1.8368, + "step": 3100 + }, + { + "epoch": 2.19, + "learning_rate": 3.6197580690556233e-06, + "loss": 1.819, + "step": 3150 + }, + { + "epoch": 2.22, + "learning_rate": 3.335835912015e-06, + "loss": 1.8237, + "step": 3200 + }, + { + "epoch": 2.26, + "learning_rate": 3.0612653333673092e-06, + "loss": 1.8163, + "step": 3250 + }, + { + "epoch": 2.29, + "learning_rate": 2.7964316279907555e-06, + "loss": 1.8223, + "step": 3300 + }, + { + "epoch": 2.32, + "learning_rate": 2.541706427364431e-06, + "loss": 1.8375, + "step": 3350 + }, + { + "epoch": 2.36, + "learning_rate": 2.2974471780714147e-06, + "loss": 1.8157, + "step": 3400 + }, + { + "epoch": 2.39, + "learning_rate": 2.0639966402070154e-06, + "loss": 1.8172, + "step": 3450 + }, + { + "epoch": 2.43, + "learning_rate": 1.8416824063960336e-06, + "loss": 1.7918, + "step": 3500 + }, + { + "epoch": 2.46, + "learning_rate": 1.6308164420939909e-06, + "loss": 1.8149, + "step": 3550 + }, + { + "epoch": 2.5, + "learning_rate": 1.4316946478173732e-06, + "loss": 1.822, + "step": 3600 + }, + { + "epoch": 2.53, + "learning_rate": 1.2445964439172277e-06, + "loss": 1.8282, + "step": 3650 + }, + { + "epoch": 2.57, + "learning_rate": 1.069784378478781e-06, + "loss": 1.819, + "step": 3700 + }, + { + "epoch": 2.6, + "learning_rate": 9.075037588972901e-07, + "loss": 1.8158, + "step": 3750 + }, + { + "epoch": 2.64, + "learning_rate": 7.579823076471216e-07, + "loss": 1.8106, + "step": 3800 + }, + { + "epoch": 2.67, + "learning_rate": 6.214298427270948e-07, + "loss": 1.8179, + "step": 3850 + }, + { + "epoch": 2.71, + "learning_rate": 4.980379832305604e-07, + "loss": 1.8284, + "step": 3900 + }, + { + "epoch": 2.74, + "learning_rate": 3.8797988045329635e-07, + "loss": 1.8309, + "step": 3950 + }, + { + "epoch": 2.78, + "learning_rate": 2.914099749166244e-07, + "loss": 1.8219, + "step": 4000 + }, + { + "epoch": 2.81, + "learning_rate": 2.08463779646646e-07, + "loss": 1.8182, + "step": 4050 + }, + { + "epoch": 2.85, + "learning_rate": 1.3925769001376876e-07, + "loss": 1.8142, + "step": 4100 + }, + { + "epoch": 2.88, + "learning_rate": 8.388882039931934e-08, + "loss": 1.8152, + "step": 4150 + }, + { + "epoch": 2.91, + "learning_rate": 4.243486791847829e-08, + "loss": 1.8172, + "step": 4200 + }, + { + "epoch": 2.95, + "learning_rate": 1.495400339075359e-08, + "loss": 1.8079, + "step": 4250 + }, + { + "epoch": 2.98, + "learning_rate": 1.4847897109848685e-09, + "loss": 1.8273, + "step": 4300 + }, + { + "epoch": 3.0, + "step": 4323, + "total_flos": 1.477332496684155e+17, + "train_loss": 1.9787378639858297, + "train_runtime": 6196.198, + "train_samples_per_second": 44.65, + "train_steps_per_second": 0.698 + } + ], + "logging_steps": 50, + "max_steps": 4323, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 1200, + "total_flos": 1.477332496684155e+17, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +} diff --git a/training_args.bin b/training_args.bin new file mode 100644 index 0000000..ef643ea --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3d9eb31f4925f42db4f735ce257100535d1ebbd738be28253c47b4fb08ae8ca1 +size 4283