From 94bf1cf9c22a1e1877931702401d30446633c21d Mon Sep 17 00:00:00 2001 From: ModelHub XC Date: Sun, 1 Nov 2026 21:23:12 +0800 Subject: [PATCH] =?UTF-8?q?=E5=88=9D=E5=A7=8B=E5=8C=96=E9=A1=B9=E7=9B=AE?= =?UTF-8?q?=EF=BC=8C=E7=94=B1ModelHub=20XC=E7=A4=BE=E5=8C=BA=E6=8F=90?= =?UTF-8?q?=E4=BE=9B=E6=A8=A1=E5=9E=8B?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Model: QuantFactory/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct-GGUF Source: Original Platform --- .gitattributes | 49 +++++++ ...-Llama-3-8B-Dolfin-v0.2-Instruct.Q2_K.gguf | 3 + ...lama-3-8B-Dolfin-v0.2-Instruct.Q3_K_L.gguf | 3 + ...lama-3-8B-Dolfin-v0.2-Instruct.Q3_K_M.gguf | 3 + ...lama-3-8B-Dolfin-v0.2-Instruct.Q3_K_S.gguf | 3 + ...-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_0.gguf | 3 + ...-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_1.gguf | 3 + ...lama-3-8B-Dolfin-v0.2-Instruct.Q4_K_M.gguf | 3 + ...lama-3-8B-Dolfin-v0.2-Instruct.Q4_K_S.gguf | 3 + ...-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_0.gguf | 3 + ...-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_1.gguf | 3 + ...lama-3-8B-Dolfin-v0.2-Instruct.Q5_K_M.gguf | 3 + ...lama-3-8B-Dolfin-v0.2-Instruct.Q5_K_S.gguf | 3 + ...-Llama-3-8B-Dolfin-v0.2-Instruct.Q6_K.gguf | 3 + ...-Llama-3-8B-Dolfin-v0.2-Instruct.Q8_0.gguf | 3 + README.md | 134 ++++++++++++++++++ configuration.json | 1 + 17 files changed, 226 insertions(+) create mode 100644 .gitattributes create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q2_K.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_L.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_M.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_S.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_0.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_1.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_M.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_S.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_0.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_1.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_M.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_S.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q6_K.gguf create mode 100644 ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q8_0.gguf create mode 100644 README.md create mode 100644 configuration.json diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..7a93679 --- /dev/null +++ b/.gitattributes @@ -0,0 +1,49 @@ +*.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 +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_1.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_0.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q8_0.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q6_K.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_0.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_1.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q2_K.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text +ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q2_K.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q2_K.gguf new file mode 100644 index 0000000..e9f7988 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q2_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a0d9fd319a01c2b2001e7769145392641fb4a36af8257098cbdcda9bffc5fd3e +size 3179131168 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_L.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_L.gguf new file mode 100644 index 0000000..92615b1 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_L.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:060506620bc8ff2bd076e7becdc4f46a18c52b81d3ff1a372c7e85582e75f2ef +size 4321956128 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_M.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_M.gguf new file mode 100644 index 0000000..efc6e0c --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fc0954ad2eeaf9fa020bf3959bf23f3e95ac395e0c636142a3db10ad9a9b0a1f +size 4018917664 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_S.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_S.gguf new file mode 100644 index 0000000..e078d31 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q3_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:258ec18c043b6b7f6a5766de161715c45d5d0b32ef6f8ce169549bc654bc531b +size 3664498976 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_0.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_0.gguf new file mode 100644 index 0000000..56e5905 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59dd087bd9ff33e507e6f95f70d91749ca46f154411fc57f2886cfaea98f7717 +size 4661211424 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_1.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_1.gguf new file mode 100644 index 0000000..9579671 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c8058113939c025fee8409c27bb4ee637c24b6028be869968a20e42c300065a +size 5130252576 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_M.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_M.gguf new file mode 100644 index 0000000..c6587ce --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:217d24339fa7dabfe44df397cda4aab41403c1a7c7736ff95752eb608972fa95 +size 4920733984 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_S.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_S.gguf new file mode 100644 index 0000000..a343aab --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q4_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9ad95b2d17c930b6a2d470b3dbc52b324e582cbe8e2e8ce6eefe1c34055f403 +size 4692668704 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_0.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_0.gguf new file mode 100644 index 0000000..c7711b9 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:69844b914ee64f6e746804d628bc6e1c106d51d64bfa508b6cd590e06203564f +size 5599293728 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_1.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_1.gguf new file mode 100644 index 0000000..248a8e6 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_1.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:13d3ff4218297426cbb9060b3826809f47a8762ed3a12391561649b92eb9f284 +size 6068334880 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_M.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_M.gguf new file mode 100644 index 0000000..f20f67b --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_M.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9cc67fb2b98775ac20d88e996d4f2ab16566cac41f1aa7b7c8b4a17b1a752aa +size 5732987168 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_S.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_S.gguf new file mode 100644 index 0000000..9bc4339 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q5_K_S.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:35588cac33b326c7b058e83423f89c53d65e917cc7e56cd188504da7cd02a520 +size 5599293728 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q6_K.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q6_K.gguf new file mode 100644 index 0000000..4ac8c88 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q6_K.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91a8084198ca09241f532480ee4d3f3b20d93fd30b98199a355205b7014e080a +size 6596006176 diff --git a/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q8_0.gguf b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q8_0.gguf new file mode 100644 index 0000000..04cd215 --- /dev/null +++ b/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct.Q8_0.gguf @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f499b48161091a8aa5bbb670b2ec22d9a52f547912e735339a8f76b7973be02b +size 8540770592 diff --git a/README.md b/README.md new file mode 100644 index 0000000..3eeabdd --- /dev/null +++ b/README.md @@ -0,0 +1,134 @@ +--- +license: llama3 +pipeline_tag: text-generation +base_model: OwenArli/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct +--- + +# QuantFactory/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct-GGUF +This is quantized version of [OwenArli/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct](https://huggingface.co/OwenArli/ArliAI-Llama-3-8B-Dolfin-v0.2-Instruct) created using llama.cpp + +# Model Description +Based on Meta-Llama-3-8b-Instruct, and is governed by Meta Llama 3 License agreement: +https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct + +v0.2 version with better improved dolphin based dataset but only 150K for testing instead of the full 850K. Doesn't seem to work that well so I will need to add the rest of the dataset. + +We are happy for anyone to try it out and give some feedback. + + +Training: +- 4096 sequence length, while the base model is 8192 sequence length. From testing it still performs the same 8192 context just fine. +- Trained on a modified and improved version of Cognitive Computations Eric Hartford's Dolphin dataset. https://huggingface.co/datasets/cognitivecomputations/dolphin +- Training duration is around 1 day on 2x RTX3090 on our own machine, using 4-bit loading and Qlora 64-rank 128-alpha resulting in ~2% trainable weights. + + +The goal for this model is to have the model less-censored and great at general tasks like the previous dolphin based models by Eric Hartford. + + +Instruct format: +``` +<|begin_of_text|><|start_header_id|>system<|end_header_id|> + +{{ system_prompt }}<|eot_id|><|start_header_id|>user<|end_header_id|> + +{{ user_message_1 }}<|eot_id|><|start_header_id|>assistant<|end_header_id|> + +{{ model_answer_1 }}<|eot_id|><|start_header_id|>user<|end_header_id|> + +{{ user_message_2 }}<|eot_id|><|start_header_id|>assistant<|end_header_id|> +``` + + +Quants: + + + +[Built with Axolotl](https://github.com/OpenAccess-AI-Collective/axolotl) + +Axolotl Config: +``` +base_model: /home/owen/models/Meta-Llama-3-8B-Instruct +model_type: LlamaForCausalLM +tokenizer_type: AutoTokenizer + +train_on_inputs: false +group_by_length: false +load_in_8bit: false +load_in_4bit: true +strict: false +sequence_len: 4096 +bf16: true +fp16: false +tf32: false +flash_attention: true + +# Data +datasets: + - path: /home/owen/datasets/cleaned-dolphin201-sharegpt2-uuid-improved.jsonl + type: + field_instruction: input + field_output: output + format: "<|start_header_id|>user<|end_header_id|>\n\n{instruction}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n" + no_input_format: "<|start_header_id|>user<|end_header_id|>\n\n{instruction}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n" + +warmup_steps: 10 +dataset_prepared_path: ./last_run_prepared + +# Iterations +num_epochs: 1 +saves_per_epoch: 4 + +# Evaluation +val_set_size: 0.01 +eval_table_size: +eval_table_max_new_tokens: +eval_sample_packing: false +evals_per_epoch: 4 + +# LoRA +output_dir: ./qlora-out +adapter: qlora +lora_model_dir: +lora_r: 64 +lora_alpha: 128 +lora_dropout: 0.05 +lora_target_linear: true +lora_fan_in_fan_out: +lora_target_modules: +save_safetensors: true + +# Sampling +sample_packing: true +pad_to_sequence_len: true + +# Batching +gradient_accumulation_steps: 32 +micro_batch_size: 2 +gradient_checkpointing: true +gradient_checkpointing_kwargs: + use_reentrant: true + +# wandb +wandb_mode: # "offline" to save run metadata locally and not sync to the server, "disabled" to turn off wandb +wandb_project: llama-3-8b-instruct-dolphin-q +wandb_entity: # A wandb Team name if using a Team +wandb_watch: +wandb_name: 64-128-4096-1ep-v0.2 +wandb_run_id: # Set the ID of your wandb run +wandb_log_model: # "checkpoint" to log model to wandb Artifacts every `save_steps` or "end" to log only at the end of training + +# Optimizer +optimizer: paged_adamw_8bit +lr_scheduler: cosine +learning_rate: 0.0002 + +# Misc +early_stopping_patience: +resume_from_checkpoint: +logging_steps: 1 +debug: +deepspeed: /home/owen/axolotl/deepspeed_configs/zero3_bf16.json +weight_decay: 0.1 +special_tokens: + pad_token: <|end_of_text|> +``` \ No newline at end of file diff --git a/configuration.json b/configuration.json new file mode 100644 index 0000000..9e26dfe --- /dev/null +++ b/configuration.json @@ -0,0 +1 @@ +{} \ No newline at end of file