返回模型

说明文档

启用键值缓存的 Llama 3.2 3B Instruct ONNX AWQ（4位）格式

模型创建者：Meta-Llama
原始模型：Meta-Llama Llama 3.2 3B Instruct

描述

本仓库包含由 Esperanto Technologies 转换的 Llama 3.2 3B Instruct ONNX 文件。该模型采用 AWQ 量化的 4 位格式，并启用了 KVC（键值缓存）。

关于 AWQ

AWQ 是一种高效、准确且极速的低比特权重量化方法，目前支持 4 位量化。与 GPTQ 相比，它在基于 Transformers 的推理中速度更快，且质量等同于或优于最常用的 GPTQ 设置。更多信息请参阅：AutoAWQ

如何下载 ONNX 模型和权重文件

获取模型最简单的方法是克隆整个仓库。或者，您可以使用 huggingface-hub Python 库下载文件。

pip3 install huggingface-hub>=0.17.1

然后，您可以使用如下命令高速下载任意单个模型文件到当前目录：

huggingface-cli download Esperanto/llama-3.2-3B-Instruct-kvc-AWQ-int4-onnx --local-dir llama-3.2-3B-Instruct-kvc-AWQ-int4-onnx --local-dir-use-symlinks False

有关使用 huggingface-cli 下载的更多文档，请参阅：HF -> Hub Python Library -> Download files -> Download from the CLI。

如何使用 ONNXRuntime 从 Python 代码运行

该模型可以轻松地在 CPU 上使用 ONNXRuntime 运行。

首先安装依赖包

pip3 install onnx==1.16.1
pip3 install onnxruntime==1.17.1

示例代码：使用此模型生成文本

我们定义使用贪婪解码的循环：

import numpy as np
import onnxruntime
import onnx
from transformers import AutoTokenizer

def generate_text(model_path, prompt, tokenizer, max_gen_tokens, total_sequence, window, context):
    model = onnx.load(model_path)

    #we create the inputs for the first iteration
    input_tensor = tokenizer(prompt, return_tensors=\"pt\")
    prompt_size = len(input_tensor['input_ids'][0])
    actual_input = input_tensor['input_ids']
    if prompt_size < window:
        actual_input = np.concatenate((tokenizer.bos_token_id*np.ones([1, window - prompt_size], dtype = 'int64'),
                                       actual_input), axis=1)
    if prompt_size + max_gen_tokens > total_sequence:
        print(\"ERROR: Longer total sequence is needed!\")
        return
    first_attention = np.concatenate((np.zeros([1, total_sequence - window], dtype = 'int64'),
                                      np.ones((1, window), dtype = 'int64')), axis=1)
    max_gen_tokens += prompt_size #we need to generate on top of parsing the prompt
    inputs_names =[node.name for node in model.graph.input]
    output_names =[node.name for node in model.graph.output]
    n_heads = 8 #gqa-heads of the kvc
    inputs_dict = {}
    inputs_dict['input_ids'] = actual_input[:, :window].reshape(1, window).numpy()
    inputs_dict['attention_mask'] = first_attention
    index_pos = sum(first_attention[0])
    inputs_dict['position_ids'] = np.concatenate((np.zeros([1, total_sequence - index_pos], dtype = 'int64'), np.arange(index_pos, dtype = 'int64').reshape(1, index_pos)), axis=1)
    inputs_dict['tree_attention'] = np.triu(-65504*np.ones(total_sequence), k= 1).astype('float16').reshape(1, 1, total_sequence, total_sequence)
    for name in inputs_names:
        if name == 'input_ids' or name == 'attention_mask' or name == 'position_ids' or name == 'tree_attention': continue
        inputs_dict[name] = np.zeros([1, n_heads, context-window, 128], dtype=\"float16\")
    index = 0
    new_token = np.array([10])
    next_index = window
    old_j = 0
    total_input = actual_input.numpy()

    rt_session = onnxruntime.InferenceSession(model_path)
    ## We run the inferences
    while next_index < max_gen_tokens:
        if new_token.any() == tokenizer.eos_token_id:
            break
        #inference
        output = rt_session.run(output_names, inputs_dict)
        outs_dictionary = {name: content for (name, content) in zip (output_names, output)}
        #we prepare the inputs for the next inference
        for name in inputs_names:
            if name == 'input_ids':
                old_j = next_index
                if next_index < prompt_size:
                    if prompt_size - next_index >= window: next_index += window
                    else: next_index = prompt_size 
                    j = next_index - window
                else:
                    next_index +=1
                    j = next_index - window
                    new_token = outs_dictionary['logits'].argmax(-1).reshape(1, window)
                    total_input = np.concatenate((total_input, new_token[: , -1:]), axis = 1)
                inputs_dict['input_ids']= total_input[:, j:next_index].reshape(1, window)
            elif name == 'attention_mask':
                inputs_dict['attention_mask'] = np.concatenate((np.zeros((1, total_sequence-next_index), dtype = 'int64'), np.ones((1, next_index), dtype = 'int64')), axis=1)
            elif name == 'position_ids':
                inputs_dict['position_ids'] = np.concatenate((np.zeros([1, total_sequence - next_index], dtype = 'int64'), np.arange(next_index, dtype = 'int64').reshape(1, next_index)), axis=1)
            elif name == 'tree_attention': continue
            else:
                old_name = name.replace(\"past_key_values\", \"present\")
                inputs_dict[name] = outs_dictionary[old_name][:, :, next_index-old_j:context-window+(next_index - old_j), :]

    answer = tokenizer.decode(total_input[0], skip_special_tokens=True, clean_up_tokenization_spaces=False)
    return answer

现在我们运行推理：

tokenizer = AutoTokenizer.from_pretrained(\"Esperanto/llama-3.2-3B-Instruct-kvc-AWQ-int4-onnx\")
model_path = \"llama-3.2-3B-Instruct-kvc-AWQ-int4-onnx/model.onnx\"

max_gen_tokens = 20    #number of tokens we want tog eneral
total_sequence = 128   #total sequence_length
context = 1024         #the context to extend the kvc
window = 16            #number of tokens we want to parse at the time
messages = [
    {\"role\": \"system\", \"content\": \"You are a pirate chatbot who always responds in pirate speak!\"},
    {\"role\": \"user\", \"content\": \"Who are you?\"},
]

prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)

generated = generate_text(model_path, prompt, tokenizer, max_gen_tokens, total_sequence, window, context)
print(generated)

Esperanto/llama-3.2-3B-Instruct-kvc-AWQ-int4-onnx

作者 Esperanto

text-generation

↓ 0 ♥ 0

创建时间: 2024-10-24 13:41:23+00:00

更新时间: 2024-12-12 09:47:47+00:00

在 Hugging Face 上查看

文件 (655)

.gitattributes

README.md

config.json

model.embed_tokens.weight

model.layers.0.input_layernorm.weight

model.layers.0.mlp.down_proj.qweight

model.layers.0.mlp.down_proj.qzeros

model.layers.0.mlp.down_proj.scales

model.layers.0.mlp.gate_proj.qweight

model.layers.0.mlp.gate_proj.qzeros

model.layers.0.mlp.gate_proj.scales

model.layers.0.mlp.up_proj.qweight

model.layers.0.mlp.up_proj.qzeros

model.layers.0.mlp.up_proj.scales

model.layers.0.post_attention_layernorm.weight

model.layers.0.self_attn.k_proj.qweight

model.layers.0.self_attn.k_proj.qzeros

model.layers.0.self_attn.k_proj.scales

model.layers.0.self_attn.o_proj.qweight

model.layers.0.self_attn.o_proj.qzeros

model.layers.0.self_attn.o_proj.scales

model.layers.0.self_attn.q_proj.qweight

model.layers.0.self_attn.q_proj.qzeros

model.layers.0.self_attn.q_proj.scales

model.layers.0.self_attn.v_proj.qweight

model.layers.0.self_attn.v_proj.qzeros

model.layers.0.self_attn.v_proj.scales

model.layers.1.input_layernorm.weight

model.layers.1.mlp.down_proj.qweight

model.layers.1.mlp.down_proj.qzeros

model.layers.1.mlp.down_proj.scales

model.layers.1.mlp.gate_proj.qweight

model.layers.1.mlp.gate_proj.qzeros

model.layers.1.mlp.gate_proj.scales

model.layers.1.mlp.up_proj.qweight

model.layers.1.mlp.up_proj.qzeros

model.layers.1.mlp.up_proj.scales

model.layers.1.post_attention_layernorm.weight

model.layers.1.self_attn.k_proj.qweight

model.layers.1.self_attn.k_proj.qzeros

model.layers.1.self_attn.k_proj.scales

model.layers.1.self_attn.o_proj.qweight

model.layers.1.self_attn.o_proj.qzeros

model.layers.1.self_attn.o_proj.scales

model.layers.1.self_attn.q_proj.qweight

model.layers.1.self_attn.q_proj.qzeros

model.layers.1.self_attn.q_proj.scales

model.layers.1.self_attn.v_proj.qweight

model.layers.1.self_attn.v_proj.qzeros

model.layers.1.self_attn.v_proj.scales

model.layers.10.input_layernorm.weight

model.layers.10.mlp.down_proj.qweight

model.layers.10.mlp.down_proj.qzeros

model.layers.10.mlp.down_proj.scales

model.layers.10.mlp.gate_proj.qweight

model.layers.10.mlp.gate_proj.qzeros

model.layers.10.mlp.gate_proj.scales

model.layers.10.mlp.up_proj.qweight

model.layers.10.mlp.up_proj.qzeros

model.layers.10.mlp.up_proj.scales

model.layers.10.post_attention_layernorm.weight

model.layers.10.self_attn.k_proj.qweight

model.layers.10.self_attn.k_proj.qzeros

model.layers.10.self_attn.k_proj.scales

model.layers.10.self_attn.o_proj.qweight

model.layers.10.self_attn.o_proj.qzeros

model.layers.10.self_attn.o_proj.scales

model.layers.10.self_attn.q_proj.qweight

model.layers.10.self_attn.q_proj.qzeros

model.layers.10.self_attn.q_proj.scales

model.layers.10.self_attn.v_proj.qweight

model.layers.10.self_attn.v_proj.qzeros

model.layers.10.self_attn.v_proj.scales

model.layers.11.input_layernorm.weight

model.layers.11.mlp.down_proj.qweight

model.layers.11.mlp.down_proj.qzeros

model.layers.11.mlp.down_proj.scales

model.layers.11.mlp.gate_proj.qweight

model.layers.11.mlp.gate_proj.qzeros

model.layers.11.mlp.gate_proj.scales

model.layers.11.mlp.up_proj.qweight

model.layers.11.mlp.up_proj.qzeros

model.layers.11.mlp.up_proj.scales

model.layers.11.post_attention_layernorm.weight

model.layers.11.self_attn.k_proj.qweight

model.layers.11.self_attn.k_proj.qzeros

model.layers.11.self_attn.k_proj.scales

model.layers.11.self_attn.o_proj.qweight

model.layers.11.self_attn.o_proj.qzeros

model.layers.11.self_attn.o_proj.scales

model.layers.11.self_attn.q_proj.qweight

model.layers.11.self_attn.q_proj.qzeros

model.layers.11.self_attn.q_proj.scales

model.layers.11.self_attn.v_proj.qweight

model.layers.11.self_attn.v_proj.qzeros

model.layers.11.self_attn.v_proj.scales

model.layers.12.input_layernorm.weight

model.layers.12.mlp.down_proj.qweight

model.layers.12.mlp.down_proj.qzeros

model.layers.12.mlp.down_proj.scales

model.layers.12.mlp.gate_proj.qweight

model.layers.12.mlp.gate_proj.qzeros

model.layers.12.mlp.gate_proj.scales

model.layers.12.mlp.up_proj.qweight

model.layers.12.mlp.up_proj.qzeros

model.layers.12.mlp.up_proj.scales

model.layers.12.post_attention_layernorm.weight

model.layers.12.self_attn.k_proj.qweight

model.layers.12.self_attn.k_proj.qzeros

model.layers.12.self_attn.k_proj.scales

model.layers.12.self_attn.o_proj.qweight

model.layers.12.self_attn.o_proj.qzeros

model.layers.12.self_attn.o_proj.scales

model.layers.12.self_attn.q_proj.qweight

model.layers.12.self_attn.q_proj.qzeros

model.layers.12.self_attn.q_proj.scales

model.layers.12.self_attn.v_proj.qweight

model.layers.12.self_attn.v_proj.qzeros

model.layers.12.self_attn.v_proj.scales

model.layers.13.input_layernorm.weight

model.layers.13.mlp.down_proj.qweight

model.layers.13.mlp.down_proj.qzeros

model.layers.13.mlp.down_proj.scales

model.layers.13.mlp.gate_proj.qweight

model.layers.13.mlp.gate_proj.qzeros

model.layers.13.mlp.gate_proj.scales

model.layers.13.mlp.up_proj.qweight

model.layers.13.mlp.up_proj.qzeros

model.layers.13.mlp.up_proj.scales

model.layers.13.post_attention_layernorm.weight

model.layers.13.self_attn.k_proj.qweight

model.layers.13.self_attn.k_proj.qzeros

model.layers.13.self_attn.k_proj.scales

model.layers.13.self_attn.o_proj.qweight

model.layers.13.self_attn.o_proj.qzeros

model.layers.13.self_attn.o_proj.scales

model.layers.13.self_attn.q_proj.qweight

model.layers.13.self_attn.q_proj.qzeros

model.layers.13.self_attn.q_proj.scales

model.layers.13.self_attn.v_proj.qweight

model.layers.13.self_attn.v_proj.qzeros

model.layers.13.self_attn.v_proj.scales

model.layers.14.input_layernorm.weight

model.layers.14.mlp.down_proj.qweight

model.layers.14.mlp.down_proj.qzeros

model.layers.14.mlp.down_proj.scales

model.layers.14.mlp.gate_proj.qweight

model.layers.14.mlp.gate_proj.qzeros

model.layers.14.mlp.gate_proj.scales

model.layers.14.mlp.up_proj.qweight

model.layers.14.mlp.up_proj.qzeros

model.layers.14.mlp.up_proj.scales

model.layers.14.post_attention_layernorm.weight

model.layers.14.self_attn.k_proj.qweight

model.layers.14.self_attn.k_proj.qzeros

model.layers.14.self_attn.k_proj.scales

model.layers.14.self_attn.o_proj.qweight

model.layers.14.self_attn.o_proj.qzeros

model.layers.14.self_attn.o_proj.scales

model.layers.14.self_attn.q_proj.qweight

model.layers.14.self_attn.q_proj.qzeros

model.layers.14.self_attn.q_proj.scales

model.layers.14.self_attn.v_proj.qweight

model.layers.14.self_attn.v_proj.qzeros

model.layers.14.self_attn.v_proj.scales

model.layers.15.input_layernorm.weight

model.layers.15.mlp.down_proj.qweight

model.layers.15.mlp.down_proj.qzeros

model.layers.15.mlp.down_proj.scales

model.layers.15.mlp.gate_proj.qweight

model.layers.15.mlp.gate_proj.qzeros

model.layers.15.mlp.gate_proj.scales

model.layers.15.mlp.up_proj.qweight

model.layers.15.mlp.up_proj.qzeros

model.layers.15.mlp.up_proj.scales

model.layers.15.post_attention_layernorm.weight

model.layers.15.self_attn.k_proj.qweight

model.layers.15.self_attn.k_proj.qzeros

model.layers.15.self_attn.k_proj.scales

model.layers.15.self_attn.o_proj.qweight

model.layers.15.self_attn.o_proj.qzeros

model.layers.15.self_attn.o_proj.scales

model.layers.15.self_attn.q_proj.qweight

model.layers.15.self_attn.q_proj.qzeros

model.layers.15.self_attn.q_proj.scales

model.layers.15.self_attn.v_proj.qweight

model.layers.15.self_attn.v_proj.qzeros

model.layers.15.self_attn.v_proj.scales

model.layers.16.input_layernorm.weight

model.layers.16.mlp.down_proj.qweight

model.layers.16.mlp.down_proj.qzeros

model.layers.16.mlp.down_proj.scales

model.layers.16.mlp.gate_proj.qweight

model.layers.16.mlp.gate_proj.qzeros

model.layers.16.mlp.gate_proj.scales

model.layers.16.mlp.up_proj.qweight

model.layers.16.mlp.up_proj.qzeros

model.layers.16.mlp.up_proj.scales

model.layers.16.post_attention_layernorm.weight

model.layers.16.self_attn.k_proj.qweight

model.layers.16.self_attn.k_proj.qzeros

model.layers.16.self_attn.k_proj.scales

model.layers.16.self_attn.o_proj.qweight

model.layers.16.self_attn.o_proj.qzeros

model.layers.16.self_attn.o_proj.scales

model.layers.16.self_attn.q_proj.qweight

model.layers.16.self_attn.q_proj.qzeros

model.layers.16.self_attn.q_proj.scales

model.layers.16.self_attn.v_proj.qweight

model.layers.16.self_attn.v_proj.qzeros

model.layers.16.self_attn.v_proj.scales

model.layers.17.input_layernorm.weight

model.layers.17.mlp.down_proj.qweight

model.layers.17.mlp.down_proj.qzeros

model.layers.17.mlp.down_proj.scales

model.layers.17.mlp.gate_proj.qweight

model.layers.17.mlp.gate_proj.qzeros

model.layers.17.mlp.gate_proj.scales

model.layers.17.mlp.up_proj.qweight

model.layers.17.mlp.up_proj.qzeros

model.layers.17.mlp.up_proj.scales

model.layers.17.post_attention_layernorm.weight

model.layers.17.self_attn.k_proj.qweight

model.layers.17.self_attn.k_proj.qzeros

model.layers.17.self_attn.k_proj.scales

model.layers.17.self_attn.o_proj.qweight

model.layers.17.self_attn.o_proj.qzeros

model.layers.17.self_attn.o_proj.scales

model.layers.17.self_attn.q_proj.qweight

model.layers.17.self_attn.q_proj.qzeros

model.layers.17.self_attn.q_proj.scales

model.layers.17.self_attn.v_proj.qweight

model.layers.17.self_attn.v_proj.qzeros

model.layers.17.self_attn.v_proj.scales

model.layers.18.input_layernorm.weight

model.layers.18.mlp.down_proj.qweight

model.layers.18.mlp.down_proj.qzeros

model.layers.18.mlp.down_proj.scales

model.layers.18.mlp.gate_proj.qweight

model.layers.18.mlp.gate_proj.qzeros

model.layers.18.mlp.gate_proj.scales

model.layers.18.mlp.up_proj.qweight

model.layers.18.mlp.up_proj.qzeros

model.layers.18.mlp.up_proj.scales

model.layers.18.post_attention_layernorm.weight

model.layers.18.self_attn.k_proj.qweight

model.layers.18.self_attn.k_proj.qzeros

model.layers.18.self_attn.k_proj.scales

model.layers.18.self_attn.o_proj.qweight

model.layers.18.self_attn.o_proj.qzeros

model.layers.18.self_attn.o_proj.scales

model.layers.18.self_attn.q_proj.qweight

model.layers.18.self_attn.q_proj.qzeros

model.layers.18.self_attn.q_proj.scales

model.layers.18.self_attn.v_proj.qweight

model.layers.18.self_attn.v_proj.qzeros

model.layers.18.self_attn.v_proj.scales

model.layers.19.input_layernorm.weight

model.layers.19.mlp.down_proj.qweight

model.layers.19.mlp.down_proj.qzeros

model.layers.19.mlp.down_proj.scales

model.layers.19.mlp.gate_proj.qweight

model.layers.19.mlp.gate_proj.qzeros

model.layers.19.mlp.gate_proj.scales

model.layers.19.mlp.up_proj.qweight

model.layers.19.mlp.up_proj.qzeros

model.layers.19.mlp.up_proj.scales

model.layers.19.post_attention_layernorm.weight

model.layers.19.self_attn.k_proj.qweight

model.layers.19.self_attn.k_proj.qzeros

model.layers.19.self_attn.k_proj.scales

model.layers.19.self_attn.o_proj.qweight

model.layers.19.self_attn.o_proj.qzeros

model.layers.19.self_attn.o_proj.scales

model.layers.19.self_attn.q_proj.qweight

model.layers.19.self_attn.q_proj.qzeros

model.layers.19.self_attn.q_proj.scales

model.layers.19.self_attn.v_proj.qweight

model.layers.19.self_attn.v_proj.qzeros

model.layers.19.self_attn.v_proj.scales

model.layers.2.input_layernorm.weight

model.layers.2.mlp.down_proj.qweight

model.layers.2.mlp.down_proj.qzeros

model.layers.2.mlp.down_proj.scales

model.layers.2.mlp.gate_proj.qweight

model.layers.2.mlp.gate_proj.qzeros

model.layers.2.mlp.gate_proj.scales

model.layers.2.mlp.up_proj.qweight

model.layers.2.mlp.up_proj.qzeros

model.layers.2.mlp.up_proj.scales

model.layers.2.post_attention_layernorm.weight

model.layers.2.self_attn.k_proj.qweight

model.layers.2.self_attn.k_proj.qzeros

model.layers.2.self_attn.k_proj.scales

model.layers.2.self_attn.o_proj.qweight

model.layers.2.self_attn.o_proj.qzeros

model.layers.2.self_attn.o_proj.scales

model.layers.2.self_attn.q_proj.qweight

model.layers.2.self_attn.q_proj.qzeros

model.layers.2.self_attn.q_proj.scales

model.layers.2.self_attn.v_proj.qweight

model.layers.2.self_attn.v_proj.qzeros

model.layers.2.self_attn.v_proj.scales

model.layers.20.input_layernorm.weight

model.layers.20.mlp.down_proj.qweight

model.layers.20.mlp.down_proj.qzeros

model.layers.20.mlp.down_proj.scales

model.layers.20.mlp.gate_proj.qweight

model.layers.20.mlp.gate_proj.qzeros

model.layers.20.mlp.gate_proj.scales

model.layers.20.mlp.up_proj.qweight

model.layers.20.mlp.up_proj.qzeros

model.layers.20.mlp.up_proj.scales

model.layers.20.post_attention_layernorm.weight

model.layers.20.self_attn.k_proj.qweight

model.layers.20.self_attn.k_proj.qzeros

model.layers.20.self_attn.k_proj.scales

model.layers.20.self_attn.o_proj.qweight

model.layers.20.self_attn.o_proj.qzeros

model.layers.20.self_attn.o_proj.scales

model.layers.20.self_attn.q_proj.qweight

model.layers.20.self_attn.q_proj.qzeros

model.layers.20.self_attn.q_proj.scales

model.layers.20.self_attn.v_proj.qweight

model.layers.20.self_attn.v_proj.qzeros

model.layers.20.self_attn.v_proj.scales

model.layers.21.input_layernorm.weight

model.layers.21.mlp.down_proj.qweight

model.layers.21.mlp.down_proj.qzeros

model.layers.21.mlp.down_proj.scales

model.layers.21.mlp.gate_proj.qweight

model.layers.21.mlp.gate_proj.qzeros

model.layers.21.mlp.gate_proj.scales

model.layers.21.mlp.up_proj.qweight

model.layers.21.mlp.up_proj.qzeros

model.layers.21.mlp.up_proj.scales

model.layers.21.post_attention_layernorm.weight

model.layers.21.self_attn.k_proj.qweight

model.layers.21.self_attn.k_proj.qzeros

model.layers.21.self_attn.k_proj.scales

model.layers.21.self_attn.o_proj.qweight

model.layers.21.self_attn.o_proj.qzeros

model.layers.21.self_attn.o_proj.scales

model.layers.21.self_attn.q_proj.qweight

model.layers.21.self_attn.q_proj.qzeros

model.layers.21.self_attn.q_proj.scales

model.layers.21.self_attn.v_proj.qweight

model.layers.21.self_attn.v_proj.qzeros

model.layers.21.self_attn.v_proj.scales

model.layers.22.input_layernorm.weight

model.layers.22.mlp.down_proj.qweight

model.layers.22.mlp.down_proj.qzeros

model.layers.22.mlp.down_proj.scales

model.layers.22.mlp.gate_proj.qweight

model.layers.22.mlp.gate_proj.qzeros

model.layers.22.mlp.gate_proj.scales

model.layers.22.mlp.up_proj.qweight

model.layers.22.mlp.up_proj.qzeros

model.layers.22.mlp.up_proj.scales

model.layers.22.post_attention_layernorm.weight

model.layers.22.self_attn.k_proj.qweight

model.layers.22.self_attn.k_proj.qzeros

model.layers.22.self_attn.k_proj.scales

model.layers.22.self_attn.o_proj.qweight

model.layers.22.self_attn.o_proj.qzeros

model.layers.22.self_attn.o_proj.scales

model.layers.22.self_attn.q_proj.qweight

model.layers.22.self_attn.q_proj.qzeros

model.layers.22.self_attn.q_proj.scales

model.layers.22.self_attn.v_proj.qweight

model.layers.22.self_attn.v_proj.qzeros

model.layers.22.self_attn.v_proj.scales

model.layers.23.input_layernorm.weight

model.layers.23.mlp.down_proj.qweight

model.layers.23.mlp.down_proj.qzeros

model.layers.23.mlp.down_proj.scales

model.layers.23.mlp.gate_proj.qweight

model.layers.23.mlp.gate_proj.qzeros

model.layers.23.mlp.gate_proj.scales

model.layers.23.mlp.up_proj.qweight

model.layers.23.mlp.up_proj.qzeros

model.layers.23.mlp.up_proj.scales

model.layers.23.post_attention_layernorm.weight

model.layers.23.self_attn.k_proj.qweight

model.layers.23.self_attn.k_proj.qzeros

model.layers.23.self_attn.k_proj.scales

model.layers.23.self_attn.o_proj.qweight

model.layers.23.self_attn.o_proj.qzeros

model.layers.23.self_attn.o_proj.scales

model.layers.23.self_attn.q_proj.qweight

model.layers.23.self_attn.q_proj.qzeros

model.layers.23.self_attn.q_proj.scales

model.layers.23.self_attn.v_proj.qweight

model.layers.23.self_attn.v_proj.qzeros

model.layers.23.self_attn.v_proj.scales

model.layers.24.input_layernorm.weight

model.layers.24.mlp.down_proj.qweight

model.layers.24.mlp.down_proj.qzeros

model.layers.24.mlp.down_proj.scales

model.layers.24.mlp.gate_proj.qweight

model.layers.24.mlp.gate_proj.qzeros

model.layers.24.mlp.gate_proj.scales

model.layers.24.mlp.up_proj.qweight

model.layers.24.mlp.up_proj.qzeros

model.layers.24.mlp.up_proj.scales

model.layers.24.post_attention_layernorm.weight

model.layers.24.self_attn.k_proj.qweight

model.layers.24.self_attn.k_proj.qzeros

model.layers.24.self_attn.k_proj.scales

model.layers.24.self_attn.o_proj.qweight

model.layers.24.self_attn.o_proj.qzeros

model.layers.24.self_attn.o_proj.scales

model.layers.24.self_attn.q_proj.qweight

model.layers.24.self_attn.q_proj.qzeros

model.layers.24.self_attn.q_proj.scales

model.layers.24.self_attn.v_proj.qweight

model.layers.24.self_attn.v_proj.qzeros

model.layers.24.self_attn.v_proj.scales

model.layers.25.input_layernorm.weight

model.layers.25.mlp.down_proj.qweight

model.layers.25.mlp.down_proj.qzeros

model.layers.25.mlp.down_proj.scales

model.layers.25.mlp.gate_proj.qweight

model.layers.25.mlp.gate_proj.qzeros

model.layers.25.mlp.gate_proj.scales

model.layers.25.mlp.up_proj.qweight

model.layers.25.mlp.up_proj.qzeros

model.layers.25.mlp.up_proj.scales

model.layers.25.post_attention_layernorm.weight

model.layers.25.self_attn.k_proj.qweight

model.layers.25.self_attn.k_proj.qzeros

model.layers.25.self_attn.k_proj.scales

model.layers.25.self_attn.o_proj.qweight

model.layers.25.self_attn.o_proj.qzeros

model.layers.25.self_attn.o_proj.scales

model.layers.25.self_attn.q_proj.qweight

model.layers.25.self_attn.q_proj.qzeros

model.layers.25.self_attn.q_proj.scales

model.layers.25.self_attn.v_proj.qweight

model.layers.25.self_attn.v_proj.qzeros

model.layers.25.self_attn.v_proj.scales

model.layers.26.input_layernorm.weight

model.layers.26.mlp.down_proj.qweight

model.layers.26.mlp.down_proj.qzeros

model.layers.26.mlp.down_proj.scales

model.layers.26.mlp.gate_proj.qweight

model.layers.26.mlp.gate_proj.qzeros

model.layers.26.mlp.gate_proj.scales

model.layers.26.mlp.up_proj.qweight

model.layers.26.mlp.up_proj.qzeros

model.layers.26.mlp.up_proj.scales

model.layers.26.post_attention_layernorm.weight

model.layers.26.self_attn.k_proj.qweight

model.layers.26.self_attn.k_proj.qzeros

model.layers.26.self_attn.k_proj.scales

model.layers.26.self_attn.o_proj.qweight

model.layers.26.self_attn.o_proj.qzeros

model.layers.26.self_attn.o_proj.scales

model.layers.26.self_attn.q_proj.qweight

model.layers.26.self_attn.q_proj.qzeros

model.layers.26.self_attn.q_proj.scales

model.layers.26.self_attn.v_proj.qweight

model.layers.26.self_attn.v_proj.qzeros

model.layers.26.self_attn.v_proj.scales

model.layers.27.input_layernorm.weight

model.layers.27.mlp.down_proj.qweight

model.layers.27.mlp.down_proj.qzeros

model.layers.27.mlp.down_proj.scales

model.layers.27.mlp.gate_proj.qweight

model.layers.27.mlp.gate_proj.qzeros

model.layers.27.mlp.gate_proj.scales

model.layers.27.mlp.up_proj.qweight

model.layers.27.mlp.up_proj.qzeros

model.layers.27.mlp.up_proj.scales

model.layers.27.post_attention_layernorm.weight

model.layers.27.self_attn.k_proj.qweight

model.layers.27.self_attn.k_proj.qzeros

model.layers.27.self_attn.k_proj.scales

model.layers.27.self_attn.o_proj.qweight

model.layers.27.self_attn.o_proj.qzeros

model.layers.27.self_attn.o_proj.scales

model.layers.27.self_attn.q_proj.qweight

model.layers.27.self_attn.q_proj.qzeros

model.layers.27.self_attn.q_proj.scales

model.layers.27.self_attn.v_proj.qweight

model.layers.27.self_attn.v_proj.qzeros

model.layers.27.self_attn.v_proj.scales

model.layers.3.input_layernorm.weight

model.layers.3.mlp.down_proj.qweight

model.layers.3.mlp.down_proj.qzeros

model.layers.3.mlp.down_proj.scales

model.layers.3.mlp.gate_proj.qweight

model.layers.3.mlp.gate_proj.qzeros

model.layers.3.mlp.gate_proj.scales

model.layers.3.mlp.up_proj.qweight

model.layers.3.mlp.up_proj.qzeros

model.layers.3.mlp.up_proj.scales

model.layers.3.post_attention_layernorm.weight

model.layers.3.self_attn.k_proj.qweight

model.layers.3.self_attn.k_proj.qzeros

model.layers.3.self_attn.k_proj.scales

model.layers.3.self_attn.o_proj.qweight

model.layers.3.self_attn.o_proj.qzeros

model.layers.3.self_attn.o_proj.scales

model.layers.3.self_attn.q_proj.qweight

model.layers.3.self_attn.q_proj.qzeros

model.layers.3.self_attn.q_proj.scales

model.layers.3.self_attn.v_proj.qweight

model.layers.3.self_attn.v_proj.qzeros

model.layers.3.self_attn.v_proj.scales

model.layers.4.input_layernorm.weight

model.layers.4.mlp.down_proj.qweight

model.layers.4.mlp.down_proj.qzeros

model.layers.4.mlp.down_proj.scales

model.layers.4.mlp.gate_proj.qweight

model.layers.4.mlp.gate_proj.qzeros

model.layers.4.mlp.gate_proj.scales

model.layers.4.mlp.up_proj.qweight

model.layers.4.mlp.up_proj.qzeros

model.layers.4.mlp.up_proj.scales

model.layers.4.post_attention_layernorm.weight

model.layers.4.self_attn.k_proj.qweight

model.layers.4.self_attn.k_proj.qzeros

model.layers.4.self_attn.k_proj.scales

model.layers.4.self_attn.o_proj.qweight

model.layers.4.self_attn.o_proj.qzeros

model.layers.4.self_attn.o_proj.scales

model.layers.4.self_attn.q_proj.qweight

model.layers.4.self_attn.q_proj.qzeros

model.layers.4.self_attn.q_proj.scales

model.layers.4.self_attn.v_proj.qweight

model.layers.4.self_attn.v_proj.qzeros

model.layers.4.self_attn.v_proj.scales

model.layers.5.input_layernorm.weight

model.layers.5.mlp.down_proj.qweight

model.layers.5.mlp.down_proj.qzeros

model.layers.5.mlp.down_proj.scales

model.layers.5.mlp.gate_proj.qweight

model.layers.5.mlp.gate_proj.qzeros

model.layers.5.mlp.gate_proj.scales

model.layers.5.mlp.up_proj.qweight

model.layers.5.mlp.up_proj.qzeros

model.layers.5.mlp.up_proj.scales

model.layers.5.post_attention_layernorm.weight

model.layers.5.self_attn.k_proj.qweight

model.layers.5.self_attn.k_proj.qzeros

model.layers.5.self_attn.k_proj.scales

model.layers.5.self_attn.o_proj.qweight

model.layers.5.self_attn.o_proj.qzeros

model.layers.5.self_attn.o_proj.scales

model.layers.5.self_attn.q_proj.qweight

model.layers.5.self_attn.q_proj.qzeros

model.layers.5.self_attn.q_proj.scales

model.layers.5.self_attn.v_proj.qweight

model.layers.5.self_attn.v_proj.qzeros

model.layers.5.self_attn.v_proj.scales

model.layers.6.input_layernorm.weight

model.layers.6.mlp.down_proj.qweight

model.layers.6.mlp.down_proj.qzeros

model.layers.6.mlp.down_proj.scales

model.layers.6.mlp.gate_proj.qweight

model.layers.6.mlp.gate_proj.qzeros

model.layers.6.mlp.gate_proj.scales

model.layers.6.mlp.up_proj.qweight

model.layers.6.mlp.up_proj.qzeros

model.layers.6.mlp.up_proj.scales

model.layers.6.post_attention_layernorm.weight

model.layers.6.self_attn.k_proj.qweight

model.layers.6.self_attn.k_proj.qzeros

model.layers.6.self_attn.k_proj.scales

model.layers.6.self_attn.o_proj.qweight

model.layers.6.self_attn.o_proj.qzeros

model.layers.6.self_attn.o_proj.scales

model.layers.6.self_attn.q_proj.qweight

model.layers.6.self_attn.q_proj.qzeros

model.layers.6.self_attn.q_proj.scales

model.layers.6.self_attn.v_proj.qweight

model.layers.6.self_attn.v_proj.qzeros

model.layers.6.self_attn.v_proj.scales

model.layers.7.input_layernorm.weight

model.layers.7.mlp.down_proj.qweight

model.layers.7.mlp.down_proj.qzeros

model.layers.7.mlp.down_proj.scales

model.layers.7.mlp.gate_proj.qweight

model.layers.7.mlp.gate_proj.qzeros

model.layers.7.mlp.gate_proj.scales

model.layers.7.mlp.up_proj.qweight

model.layers.7.mlp.up_proj.qzeros

model.layers.7.mlp.up_proj.scales

model.layers.7.post_attention_layernorm.weight

model.layers.7.self_attn.k_proj.qweight

model.layers.7.self_attn.k_proj.qzeros

model.layers.7.self_attn.k_proj.scales

model.layers.7.self_attn.o_proj.qweight

model.layers.7.self_attn.o_proj.qzeros

model.layers.7.self_attn.o_proj.scales

model.layers.7.self_attn.q_proj.qweight

model.layers.7.self_attn.q_proj.qzeros

model.layers.7.self_attn.q_proj.scales

model.layers.7.self_attn.v_proj.qweight

model.layers.7.self_attn.v_proj.qzeros

model.layers.7.self_attn.v_proj.scales

model.layers.8.input_layernorm.weight

model.layers.8.mlp.down_proj.qweight

model.layers.8.mlp.down_proj.qzeros

model.layers.8.mlp.down_proj.scales

model.layers.8.mlp.gate_proj.qweight

model.layers.8.mlp.gate_proj.qzeros

model.layers.8.mlp.gate_proj.scales

model.layers.8.mlp.up_proj.qweight

model.layers.8.mlp.up_proj.qzeros

model.layers.8.mlp.up_proj.scales

model.layers.8.post_attention_layernorm.weight

model.layers.8.self_attn.k_proj.qweight

model.layers.8.self_attn.k_proj.qzeros

model.layers.8.self_attn.k_proj.scales

model.layers.8.self_attn.o_proj.qweight

model.layers.8.self_attn.o_proj.qzeros

model.layers.8.self_attn.o_proj.scales

model.layers.8.self_attn.q_proj.qweight

model.layers.8.self_attn.q_proj.qzeros

model.layers.8.self_attn.q_proj.scales

model.layers.8.self_attn.v_proj.qweight

model.layers.8.self_attn.v_proj.qzeros

model.layers.8.self_attn.v_proj.scales

model.layers.9.input_layernorm.weight

model.layers.9.mlp.down_proj.qweight

model.layers.9.mlp.down_proj.qzeros

model.layers.9.mlp.down_proj.scales

model.layers.9.mlp.gate_proj.qweight

model.layers.9.mlp.gate_proj.qzeros

model.layers.9.mlp.gate_proj.scales

model.layers.9.mlp.up_proj.qweight

model.layers.9.mlp.up_proj.qzeros

model.layers.9.mlp.up_proj.scales

model.layers.9.post_attention_layernorm.weight

model.layers.9.self_attn.k_proj.qweight

model.layers.9.self_attn.k_proj.qzeros

model.layers.9.self_attn.k_proj.scales

model.layers.9.self_attn.o_proj.qweight

model.layers.9.self_attn.o_proj.qzeros

model.layers.9.self_attn.o_proj.scales

model.layers.9.self_attn.q_proj.qweight

model.layers.9.self_attn.q_proj.qzeros

model.layers.9.self_attn.q_proj.scales

model.layers.9.self_attn.v_proj.qweight

model.layers.9.self_attn.v_proj.qzeros

model.layers.9.self_attn.v_proj.scales

model.norm.weight

model.onnx ONNX

onnx__MatMul_9287

special_tokens_map.json

token_id_to_str.json

tokenizer.json

tokenizer_config.json