初始化项目，由ModelHub XC社区提供模型

Model: QuixiAI/Llama-3.2-1B
Source: Original Platform

.gitattributes

@@ -0,0 +1,50 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bin.* filter=lfs diff=lfs merge=lfs -text
+*.bz2 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
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack 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
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* 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
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zstandard filter=lfs diff=lfs merge=lfs -text
+*.tfevents* filter=lfs diff=lfs merge=lfs -text
+*.db* filter=lfs diff=lfs merge=lfs -text
+*.ark* filter=lfs diff=lfs merge=lfs -text
+**/*ckpt*data* filter=lfs diff=lfs merge=lfs -text
+**/*ckpt*.meta filter=lfs diff=lfs merge=lfs -text
+**/*ckpt*.index filter=lfs diff=lfs merge=lfs -text
+ 
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.gguf* filter=lfs diff=lfs merge=lfs -text
+*.ggml filter=lfs diff=lfs merge=lfs -text
+*.llamafile* filter=lfs diff=lfs merge=lfs -text
+*.pt2 filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+
+model.safetensors filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,31 @@
+---
+base_model: meta-llama/Llama-3.2-1B-Instruct
+language:
+- en
+library_name: transformers
+license: llama3.2
+tags:
+- llama-3
+- llama
+- meta
+- facebook
+- transformers
+---
+
+Quantizing Llama-3.2-1B
+
+Eric Hartford
+
+I am creating several quants of Llama-3.1-1B for the purposes of testing vLLM Marlin.
+
+- https://huggingface.co/QuixiAI/Llama-3.2-1B
+- https://huggingface.co/QuixiAI/Llama-3.2-1B-FP8-Dynamic
+- https://huggingface.co/QuixiAI/Llama-3.2-1B-MXFP4
+- https://huggingface.co/QuixiAI/Llama-3.2-1B-NVFP4A16
+- https://huggingface.co/QuixiAI/Llama-3.2-1B-W4A16-AWQ
+- https://huggingface.co/QuixiAI/Llama-3.2-1B-W4A16-GPTQ
+- https://huggingface.co/QuixiAI/Llama-3.2-1B-W8A16-GPTQ
+
+The script I used to quant this:
+[quant.py](quant.py)
+

chat_template.jinja

@@ -0,0 +1,93 @@
+{{- bos_token }}
+{%- if custom_tools is defined %}
+    {%- set tools = custom_tools %}
+{%- endif %}
+{%- if not tools_in_user_message is defined %}
+    {%- set tools_in_user_message = true %}
+{%- endif %}
+{%- if not date_string is defined %}
+    {%- if strftime_now is defined %}
+        {%- set date_string = strftime_now("%d %b %Y") %}
+    {%- else %}
+        {%- set date_string = "26 Jul 2024" %}
+    {%- endif %}
+{%- endif %}
+{%- if not tools is defined %}
+    {%- set tools = none %}
+{%- endif %}
+
+{#- This block extracts the system message, so we can slot it into the right place. #}
+{%- if messages[0]['role'] == 'system' %}
+    {%- set system_message = messages[0]['content']|trim %}
+    {%- set messages = messages[1:] %}
+{%- else %}
+    {%- set system_message = "" %}
+{%- endif %}
+
+{#- System message #}
+{{- "<|start_header_id|>system<|end_header_id|>\n\n" }}
+{%- if tools is not none %}
+    {{- "Environment: ipython\n" }}
+{%- endif %}
+{{- "Cutting Knowledge Date: December 2023\n" }}
+{{- "Today Date: " + date_string + "\n\n" }}
+{%- if tools is not none and not tools_in_user_message %}
+    {{- "You have access to the following functions. To call a function, please respond with JSON for a function call." }}
+    {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+    {{- "Do not use variables.\n\n" }}
+    {%- for t in tools %}
+        {{- t | tojson(indent=4) }}
+        {{- "\n\n" }}
+    {%- endfor %}
+{%- endif %}
+{{- system_message }}
+{{- "<|eot_id|>" }}
+
+{#- Custom tools are passed in a user message with some extra guidance #}
+{%- if tools_in_user_message and not tools is none %}
+    {#- Extract the first user message so we can plug it in here #}
+    {%- if messages | length != 0 %}
+        {%- set first_user_message = messages[0]['content']|trim %}
+        {%- set messages = messages[1:] %}
+    {%- else %}
+        {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }}
+{%- endif %}
+    {{- '<|start_header_id|>user<|end_header_id|>\n\n' -}}
+    {{- "Given the following functions, please respond with a JSON for a function call " }}
+    {{- "with its proper arguments that best answers the given prompt.\n\n" }}
+    {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }}
+    {{- "Do not use variables.\n\n" }}
+    {%- for t in tools %}
+        {{- t | tojson(indent=4) }}
+        {{- "\n\n" }}
+    {%- endfor %}
+    {{- first_user_message + "<|eot_id|>"}}
+{%- endif %}
+
+{%- for message in messages %}
+    {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %}
+        {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' }}
+    {%- elif 'tool_calls' in message %}
+        {%- if not message.tool_calls|length == 1 %}
+            {{- raise_exception("This model only supports single tool-calls at once!") }}
+        {%- endif %}
+        {%- set tool_call = message.tool_calls[0].function %}
+        {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' -}}
+        {{- '{"name": "' + tool_call.name + '", ' }}
+        {{- '"parameters": ' }}
+        {{- tool_call.arguments | tojson }}
+        {{- "}" }}
+        {{- "<|eot_id|>" }}
+    {%- elif message.role == "tool" or message.role == "ipython" %}
+        {{- "<|start_header_id|>ipython<|end_header_id|>\n\n" }}
+        {%- if message.content is mapping or message.content is iterable %}
+            {{- message.content | tojson }}
+        {%- else %}
+            {{- message.content }}
+        {%- endif %}
+        {{- "<|eot_id|>" }}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }}
+{%- endif %}

config.json

@@ -0,0 +1,37 @@
+{
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 128000,
+  "eos_token_id": 128009,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "max_position_embeddings": 131072,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 16,
+  "num_key_value_heads": 8,
+  "pad_token_id": 128004,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": {
+    "factor": 32.0,
+    "high_freq_factor": 4.0,
+    "low_freq_factor": 1.0,
+    "original_max_position_embeddings": 8192,
+    "rope_type": "llama3"
+  },
+  "rope_theta": 500000.0,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.52.0.dev0",
+  "unsloth_fixed": true,
+  "use_cache": true,
+  "vocab_size": 128256
+}

configuration.json

@@ -0,0 +1 @@
+{"framework": "pytorch", "task": "text-generation", "allow_remote": true}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 128000,
+  "do_sample": true,
+  "eos_token_id": [
+    128001,
+    128008,
+    128009
+  ],
+  "max_length": 131072,
+  "pad_token_id": 128004,
+  "temperature": 0.6,
+  "top_p": 0.9,
+  "transformers_version": "4.52.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ff795ff6a07e6a68085d206fb84417da2f083f68391c2843cd2b8ac6df8538f
+size 2471645608

quant.py

@@ -0,0 +1,171 @@
+#!/usr/bin/env python3
+"""Convert a local BF16 model into Marlin-supported quant formats via llm-compressor."""
+
+from __future__ import annotations
+
+import gc
+import os
+import sys
+from typing import Optional
+
+import torch
+from datasets import load_dataset
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+# Allow running against the local llm-compressor checkout without installing.
+LLM_COMPRESSOR_SRC = "/home/quixi/marlin-cdna/llm-compressor/src"
+if os.path.isdir(LLM_COMPRESSOR_SRC):
+    sys.path.insert(0, LLM_COMPRESSOR_SRC)
+
+from llmcompressor import oneshot  # noqa: E402
+from llmcompressor.modifiers.awq import AWQModifier  # noqa: E402
+from llmcompressor.modifiers.quantization import (  # noqa: E402
+    GPTQModifier,
+    QuantizationModifier,
+)
+
+MODEL_PATH = "/home/quixi/models/Llama-3.2-1B"
+OUTPUT_ROOT = "/home/quixi/models"
+
+CALIB_DATASET_ID = "HuggingFaceH4/ultrachat_200k"
+CALIB_DATASET_SPLIT = "train_sft"
+NUM_CALIBRATION_SAMPLES = 128
+MAX_SEQUENCE_LENGTH = 512
+
+
+def _load_tokenized_dataset(tokenizer):
+    ds = load_dataset(
+        CALIB_DATASET_ID,
+        split=f"{CALIB_DATASET_SPLIT}[:{NUM_CALIBRATION_SAMPLES}]",
+    ).shuffle(seed=42)
+
+    def preprocess(example):
+        return {
+            "text": tokenizer.apply_chat_template(
+                example["messages"],
+                tokenize=False,
+            )
+        }
+
+    ds = ds.map(preprocess)
+
+    def tokenize(sample):
+        return tokenizer(
+            sample["text"],
+            padding=False,
+            max_length=MAX_SEQUENCE_LENGTH,
+            truncation=True,
+            add_special_tokens=False,
+        )
+
+    return ds.map(tokenize, remove_columns=ds.column_names)
+
+
+def _load_model_and_tokenizer():
+    model = AutoModelForCausalLM.from_pretrained(MODEL_PATH, dtype="auto")
+    tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
+    if torch.cuda.is_available():
+        model.to("cuda")
+    return model, tokenizer
+
+
+def _cleanup(model, tokenizer):
+    del model
+    del tokenizer
+    gc.collect()
+    if torch.cuda.is_available():
+        torch.cuda.empty_cache()
+
+
+def _run_recipe(
+    name: str,
+    recipe,
+    *,
+    save_compressed: bool,
+    use_calibration: bool,
+) -> Optional[str]:
+    print(f"\n=== Quantizing {name} ===")
+    model, tokenizer = _load_model_and_tokenizer()
+
+    oneshot_kwargs = {"model": model, "recipe": recipe}
+    if use_calibration:
+        ds = _load_tokenized_dataset(tokenizer)
+        oneshot_kwargs.update(
+            dataset=ds,
+            max_seq_length=MAX_SEQUENCE_LENGTH,
+            num_calibration_samples=NUM_CALIBRATION_SAMPLES,
+        )
+
+    oneshot(**oneshot_kwargs)
+
+    base_name = os.path.basename(MODEL_PATH.rstrip("/"))
+    save_dir = os.path.join(OUTPUT_ROOT, f"{base_name}-{name}")
+    os.makedirs(save_dir, exist_ok=True)
+
+    if save_compressed:
+        model.save_pretrained(save_dir, save_compressed=True)
+    else:
+        model.save_pretrained(save_dir)
+    tokenizer.save_pretrained(save_dir)
+
+    _cleanup(model, tokenizer)
+    return save_dir
+
+
+def main():
+    # GPTQ W4A16 (INT4 weight-only).
+    _run_recipe(
+        "W4A16-GPTQ",
+        GPTQModifier(targets="Linear", scheme="W4A16", ignore=["lm_head"]),
+        save_compressed=True,
+        use_calibration=True,
+    )
+
+    # AWQ W4A16 (INT4 weight-only).
+    _run_recipe(
+        "W4A16-AWQ",
+        AWQModifier(
+            targets=["Linear"],
+            scheme="W4A16_ASYM",
+            ignore=["lm_head"],
+            duo_scaling="both",
+        ),
+        save_compressed=True,
+        use_calibration=True,
+    )
+
+    # GPTQ W8A16 (INT8 weight-only).
+    _run_recipe(
+        "W8A16-GPTQ",
+        GPTQModifier(targets="Linear", scheme="W8A16", ignore=["lm_head"]),
+        save_compressed=True,
+        use_calibration=True,
+    )
+
+    # FP8 dynamic (W8A8-FP8).
+    _run_recipe(
+        "FP8-Dynamic",
+        QuantizationModifier(targets="Linear", scheme="FP8_DYNAMIC", ignore=["lm_head"]),
+        save_compressed=False,
+        use_calibration=False,
+    )
+
+    # NVFP4A16 (FP4 weights + FP16 activations).
+    _run_recipe(
+        "NVFP4A16",
+        QuantizationModifier(targets="Linear", scheme="NVFP4A16", ignore=["lm_head"]),
+        save_compressed=True,
+        use_calibration=False,
+    )
+
+    # MXFP4 (FP4 weights).
+    _run_recipe(
+        "MXFP4",
+        QuantizationModifier(targets="Linear", scheme="MXFP4", ignore=["lm_head"]),
+        save_compressed=True,
+        use_calibration=False,
+    )
+
+
+if __name__ == "__main__":
+    main()

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<|begin_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|eot_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|finetune_right_pad_id|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}