duyntnet/Qwen3-4B-imatrix-GGUF
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: duyntnet/Qwen3-4B-imatrix-GGUF
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-06-04 00:36:19 +08:00
commit 6c7e5e9292
29 changed files with 499 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

62
.gitattributes vendored Normal file
View File

@@ -0,0 +1,62 @@
*.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
Qwen3-4B-IQ1_S.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ1_M.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q2_K_S.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ2_S.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ2_XS.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ2_XXS.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ3_S.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q4_1.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q5_0.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q5_1.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text
Qwen3-4B-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text

3
Qwen3-4B-IQ1_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:a042e0a190ba87844f0ca17c4362cfc7ab32922e52d35e986a640fc9c974a1cb
size 1127018368

3
Qwen3-4B-IQ1_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:366422167c4d4e3213c019222ba9f1289995938b563605f3fb6369d42f8d5ac3
size 1055256448

3
Qwen3-4B-IQ2_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:5d0abebbfdf34714740e43ed307d409ef44d86d140b8ab1ccd77d1db47fc8f40
size 1512984448

3
Qwen3-4B-IQ2_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:bbba185cb5a5f50f28fd2cd11ee008bfd21d22eb0f429e1975df8a1217ed9096
size 1417301888

3
Qwen3-4B-IQ2_XS.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:fdabc565a6ca686f37e0a80f2d6608f6d2098857a1cc19136a11b87ac5047f80
size 1354100608

3
Qwen3-4B-IQ2_XXS.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:6a14dd79d3311da4b5295bb07b31b5597acd9a084c7264f6bf226e5444e97f31
size 1246621568

3
Qwen3-4B-IQ3_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:d087eafbb40e6605859a1c18559293450194cfe0e08c7c6a024624aaf5451082
size 1962896768

3
Qwen3-4B-IQ3_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:bbce76fab0940f9a68eaa79745ae005a314a10786aa020f2679593d62f1acebb
size 1899531648

3
Qwen3-4B-IQ3_XS.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:714557106064707d2edb2ac72b2844f2d3ba7ac84a5b8c4374785ef159b1f36d
size 1814375808

3
Qwen3-4B-IQ3_XXS.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:7265a1a09d92b0d13e2699592cac379c60a0c411e79a3cab1f40abe97034d57e
size 1670188928

3
Qwen3-4B-IQ4_NL.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:e99f0b49feec35e0059460fbe07434d592637231889cc84aa5dcb90c92756083
size 2381344128

3
Qwen3-4B-IQ4_XS.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:a41c4ed4114842314cf6326eaaf79e8587d4a74dc4dd2f6ba508b853d1be8a16
size 2270752128

3
Qwen3-4B-Q2_K.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:b7159aeb18cb11e0497d67229d56e802002fe3c248ea96c7d83c41685cebb05f
size 1669500288

3
Qwen3-4B-Q2_K_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:9181a9cdcdbe63d258faaa3141cc96215b1d8ec11c31cd150703258aadc68de9
size 1563454848

3
Qwen3-4B-Q3_K_L.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:11b224ba785109659d49ae787d6dd55851a60ae986e9c5aab9e123007c0cf027
size 2239786368

3
Qwen3-4B-Q3_K_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:0ebab58b726181cbc11e8d958b0a35685f8ad6e95a5d043f0964c1c880ad4fbe
size 2075618688

3
Qwen3-4B-Q3_K_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:c7af7a505c2d69550f7446422b09ce511b577d935bdf0f17a4c199ea471dbfd8
size 1886997888

3
Qwen3-4B-Q4_0.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:8738dd3ed3349b873485f2da12f241b5595f4623e089530c2f057ca72809d147
size 2375773568

3
Qwen3-4B-Q4_1.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:69ba21d65b6c502c01e62159ba54e8db1b09ca90f126f086b17b5b9b817e9d92
size 2596629888

3
Qwen3-4B-Q4_K_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:9433caa090c4b24dd81589e9248b9508a5009841462ee11529f92a2c5effec1e
size 2497281408

3
Qwen3-4B-Q4_K_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:2a905fb4839b734b1a8547cabd1f2475850c6d1aeb097766ed93df9065c2ca39
size 2383310208

3
Qwen3-4B-Q5_0.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:c262393aa75cf34b79d68e931bcc48a3f40b52c6eebf22e90f0efb0225c5827b
size 2829938048

3
Qwen3-4B-Q5_1.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:c9a5004305bc12d00499ffef264355ab34fb7f73fa83272d1f8544f0abe76ae2
size 3050794368

3
Qwen3-4B-Q5_K_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:428c551363de9d4d7e48c1335c4d0a9b3c0a5f692a82036a54347657c5db4bdf
size 2889514368

3
Qwen3-4B-Q5_K_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:0e3b86992429c6ea388127575c51211677f2ce564deb17507ced53b389e91e40
size 2823712128

3
Qwen3-4B-Q6_K.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:a52a0b80f5e1368c1e4f2d89f6c98acb97794880df7ad7f4dab8a03471bcacbc
size 3306261888

3
Qwen3-4B-Q8_0.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:579ca2480393e5970b00e944ac80b548092dc186fd55e6a858fa27790efc5664
size 4280405888

356
README.md Normal file
View File

@@ -0,0 +1,356 @@
---
license: other
language:
- en
pipeline_tag: text-generation
inference: false
tags:
- transformers
- gguf
- imatrix
- Qwen3-4B
---
Quantizations of https://huggingface.co/Qwen/Qwen3-4B
### Open source inference clients/UIs
* [llama.cpp](https://github.com/ggerganov/llama.cpp)
* [KoboldCPP](https://github.com/LostRuins/koboldcpp)
* [text-generation-webui](https://github.com/oobabooga/text-generation-webui)
* [ollama](https://github.com/ollama/ollama)
* [jan](https://github.com/janhq/jan)
* [GPT4All](https://github.com/nomic-ai/gpt4all)
### Closed source inference clients/UIs
* [LM Studio](https://lmstudio.ai/)
* [Backyard AI](https://backyard.ai/)
* More will be added...
---
# From original readme
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support, with the following key features:
- **Uniquely support of seamless switching between thinking mode** (for complex logical reasoning, math, and coding) and **non-thinking mode** (for efficient, general-purpose dialogue) **within single model**, ensuring optimal performance across various scenarios.
- **Significantly enhancement in its reasoning capabilities**, surpassing previous QwQ (in thinking mode) and Qwen2.5 instruct models (in non-thinking mode) on mathematics, code generation, and commonsense logical reasoning.
- **Superior human preference alignment**, excelling in creative writing, role-playing, multi-turn dialogues, and instruction following, to deliver a more natural, engaging, and immersive conversational experience.
- **Expertise in agent capabilities**, enabling precise integration with external tools in both thinking and unthinking modes and achieving leading performance among open-source models in complex agent-based tasks.
- **Support of 100+ languages and dialects** with strong capabilities for **multilingual instruction following** and **translation**.
## Model Overview
**Qwen3-4B** has the following features:
- Type: Causal Language Models
- Training Stage: Pretraining & Post-training
- Number of Parameters: 4.0B
- Number of Paramaters (Non-Embedding): 3.6B
- Number of Layers: 36
- Number of Attention Heads (GQA): 32 for Q and 8 for KV
- Context Length: 32,768 natively and [131,072 tokens with YaRN](#processing-long-texts).
For more details, including benchmark evaluation, hardware requirements, and inference performance, please refer to our [blog](https://qwenlm.github.io/blog/qwen3/), [GitHub](https://github.com/QwenLM/Qwen3), and [Documentation](https://qwen.readthedocs.io/en/latest/).
> [!TIP]
> If you encounter significant endless repetitions, please refer to the [Best Practices](#best-practices) section for optimal sampling parameters, and set the ``presence_penalty`` to 1.5.
## Quickstart
The code of Qwen3 has been in the latest Hugging Face `transformers` and we advise you to use the latest version of `transformers`.
With `transformers<4.51.0`, you will encounter the following error:
```
KeyError: 'qwen3'
```
The following contains a code snippet illustrating how to use the model generate content based on given inputs.
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen3-4B"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 (</think>)
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
For deployment, you can use `sglang>=0.4.6.post1` or `vllm>=0.8.4` or to create an OpenAI-compatible API endpoint:
- SGLang:
```shell
python -m sglang.launch_server --model-path Qwen/Qwen3-4B --reasoning-parser qwen3
```
- vLLM:
```shell
vllm serve Qwen/Qwen3-4B --enable-reasoning --reasoning-parser deepseek_r1
```
For local use, applications such as llama.cpp, Ollama, LMStudio, and MLX-LM have also supported Qwen3.
## Switching Between Thinking and Non-Thinking Mode
> [!TIP]
> The `enable_thinking` switch is also available in APIs created by SGLang and vLLM.
> Please refer to our documentation for [SGLang](https://qwen.readthedocs.io/en/latest/deployment/sglang.html#thinking-non-thinking-modes) and [vLLM](https://qwen.readthedocs.io/en/latest/deployment/vllm.html#thinking-non-thinking-modes) users.
### `enable_thinking=True`
By default, Qwen3 has thinking capabilities enabled, similar to QwQ-32B. This means the model will use its reasoning abilities to enhance the quality of generated responses. For example, when explicitly setting `enable_thinking=True` or leaving it as the default value in `tokenizer.apply_chat_template`, the model will engage its thinking mode.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # True is the default value for enable_thinking
)
```
In this mode, the model will generate think content wrapped in a `<think>...</think>` block, followed by the final response.
> [!NOTE]
> For thinking mode, use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0` (the default setting in `generation_config.json`). **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### `enable_thinking=False`
We provide a hard switch to strictly disable the model's thinking behavior, aligning its functionality with the previous Qwen2.5-Instruct models. This mode is particularly useful in scenarios where disabling thinking is essential for enhancing efficiency.
```python
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=False # Setting enable_thinking=False disables thinking mode
)
```
In this mode, the model will not generate any think content and will not include a `<think>...</think>` block.
> [!NOTE]
> For non-thinking mode, we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`. For more detailed guidance, please refer to the [Best Practices](#best-practices) section.
### Advanced Usage: Switching Between Thinking and Non-Thinking Modes via User Input
We provide a soft switch mechanism that allows users to dynamically control the model's behavior when `enable_thinking=True`. Specifically, you can add `/think` and `/no_think` to user prompts or system messages to switch the model's thinking mode from turn to turn. The model will follow the most recent instruction in multi-turn conversations.
Here is an example of a multi-turn conversation:
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
class QwenChatbot:
def __init__(self, model_name="Qwen/Qwen3-4B"):
self.tokenizer = AutoTokenizer.from_pretrained(model_name)
self.model = AutoModelForCausalLM.from_pretrained(model_name)
self.history = []
def generate_response(self, user_input):
messages = self.history + [{"role": "user", "content": user_input}]
text = self.tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
inputs = self.tokenizer(text, return_tensors="pt")
response_ids = self.model.generate(**inputs, max_new_tokens=32768)[0][len(inputs.input_ids[0]):].tolist()
response = self.tokenizer.decode(response_ids, skip_special_tokens=True)
# Update history
self.history.append({"role": "user", "content": user_input})
self.history.append({"role": "assistant", "content": response})
return response
# Example Usage
if __name__ == "__main__":
chatbot = QwenChatbot()
# First input (without /think or /no_think tags, thinking mode is enabled by default)
user_input_1 = "How many r's in strawberries?"
print(f"User: {user_input_1}")
response_1 = chatbot.generate_response(user_input_1)
print(f"Bot: {response_1}")
print("----------------------")
# Second input with /no_think
user_input_2 = "Then, how many r's in blueberries? /no_think"
print(f"User: {user_input_2}")
response_2 = chatbot.generate_response(user_input_2)
print(f"Bot: {response_2}")
print("----------------------")
# Third input with /think
user_input_3 = "Really? /think"
print(f"User: {user_input_3}")
response_3 = chatbot.generate_response(user_input_3)
print(f"Bot: {response_3}")
```
> [!NOTE]
> For API compatibility, when `enable_thinking=True`, regardless of whether the user uses `/think` or `/no_think`, the model will always output a block wrapped in `<think>...</think>`. However, the content inside this block may be empty if thinking is disabled.
> When `enable_thinking=False`, the soft switches are not valid. Regardless of any `/think` or `/no_think` tags input by the user, the model will not generate think content and will not include a `<think>...</think>` block.
## Agentic Use
Qwen3 excels in tool calling capabilities. We recommend using [Qwen-Agent](https://github.com/QwenLM/Qwen-Agent) to make the best use of agentic ability of Qwen3. Qwen-Agent encapsulates tool-calling templates and tool-calling parsers internally, greatly reducing coding complexity.
To define the available tools, you can use the MCP configuration file, use the integrated tool of Qwen-Agent, or integrate other tools by yourself.
```python
from qwen_agent.agents import Assistant
# Define LLM
llm_cfg = {
'model': 'Qwen3-4B',
# Use the endpoint provided by Alibaba Model Studio:
# 'model_type': 'qwen_dashscope',
# 'api_key': os.getenv('DASHSCOPE_API_KEY'),
# Use a custom endpoint compatible with OpenAI API:
'model_server': 'http://localhost:8000/v1', # api_base
'api_key': 'EMPTY',
# Other parameters:
# 'generate_cfg': {
# # Add: When the response content is `<think>this is the thought</think>this is the answer;
# # Do not add: When the response has been separated by reasoning_content and content.
# 'thought_in_content': True,
# },
}
# Define Tools
tools = [
{'mcpServers': { # You can specify the MCP configuration file
'time': {
'command': 'uvx',
'args': ['mcp-server-time', '--local-timezone=Asia/Shanghai']
},
"fetch": {
"command": "uvx",
"args": ["mcp-server-fetch"]
}
}
},
'code_interpreter', # Built-in tools
]
# Define Agent
bot = Assistant(llm=llm_cfg, function_list=tools)
# Streaming generation
messages = [{'role': 'user', 'content': 'https://qwenlm.github.io/blog/ Introduce the latest developments of Qwen'}]
for responses in bot.run(messages=messages):
pass
print(responses)
```
## Processing Long Texts
Qwen3 natively supports context lengths of up to 32,768 tokens. For conversations where the total length (including both input and output) significantly exceeds this limit, we recommend using RoPE scaling techniques to handle long texts effectively. We have validated the model's performance on context lengths of up to 131,072 tokens using the [YaRN](https://arxiv.org/abs/2309.00071) method.
YaRN is currently supported by several inference frameworks, e.g., `transformers` and `llama.cpp` for local use, `vllm` and `sglang` for deployment. In general, there are two approaches to enabling YaRN for supported frameworks:
- Modifying the model files:
In the `config.json` file, add the `rope_scaling` fields:
```json
{
...,
"rope_scaling": {
"type": "yarn",
"factor": 4.0,
"original_max_position_embeddings": 32768
}
}
```
For `llama.cpp`, you need to regenerate the GGUF file after the modification.
- Passing command line arguments:
For `vllm`, you can use
```shell
vllm serve ... --rope-scaling '{"type":"yarn","factor":4.0,"original_max_position_embeddings":32768}' --max-model-len 131072
```
For `sglang`, you can use
```shell
python -m sglang.launch_server ... --json-model-override-args '{"rope_scaling":{"type":"yarn","factor":4.0,"original_max_position_embeddings":32768}}'
```
For `llama-server` from `llama.cpp`, you can use
```shell
llama-server ... --rope-scaling yarn --rope-scale 4 --yarn-orig-ctx 32768
```
> [!IMPORTANT]
> If you encounter the following warning
> ```
> Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'original_max_position_embeddings'}
> ```
> please upgrade `transformers>=4.51.0`.
> [!NOTE]
> All the notable open-source frameworks implement static YaRN, which means the scaling factor remains constant regardless of input length, **potentially impacting performance on shorter texts.**
> We advise adding the `rope_scaling` configuration only when processing long contexts is required.
> It is also recommended to modify the `factor` as needed. For example, if the typical context length for your application is 65,536 tokens, it would be better to set `factor` as 2.0.
> [!NOTE]
> The default `max_position_embeddings` in `config.json` is set to 40,960. This allocation includes reserving 32,768 tokens for outputs and 8,192 tokens for typical prompts, which is sufficient for most scenarios involving short text processing. If the average context length does not exceed 32,768 tokens, we do not recommend enabling YaRN in this scenario, as it may potentially degrade model performance.
> [!TIP]
> The endpoint provided by Alibaba Model Studio supports dynamic YaRN by default and no extra configuration is needed.
## Best Practices
To achieve optimal performance, we recommend the following settings:
1. **Sampling Parameters**:
- For thinking mode (`enable_thinking=True`), use `Temperature=0.6`, `TopP=0.95`, `TopK=20`, and `MinP=0`. **DO NOT use greedy decoding**, as it can lead to performance degradation and endless repetitions.
- For non-thinking mode (`enable_thinking=False`), we suggest using `Temperature=0.7`, `TopP=0.8`, `TopK=20`, and `MinP=0`.
- For supported frameworks, you can adjust the `presence_penalty` parameter between 0 and 2 to reduce endless repetitions. However, using a higher value may occasionally result in language mixing and a slight decrease in model performance.
2. **Adequate Output Length**: We recommend using an output length of 32,768 tokens for most queries. For benchmarking on highly complex problems, such as those found in math and programming competitions, we suggest setting the max output length to 38,912 tokens. This provides the model with sufficient space to generate detailed and comprehensive responses, thereby enhancing its overall performance.
3. **Standardize Output Format**: We recommend using prompts to standardize model outputs when benchmarking.
- **Math Problems**: Include "Please reason step by step, and put your final answer within \boxed{}." in the prompt.
- **Multiple-Choice Questions**: Add the following JSON structure to the prompt to standardize responses: "Please show your choice in the `answer` field with only the choice letter, e.g., `"answer": "C"`."
4. **No Thinking Content in History**: In multi-turn conversations, the historical model output should only include the final output part and does not need to include the thinking content. It is implemented in the provided chat template in Jinja2. However, for frameworks that do not directly use the Jinja2 chat template, it is up to the developers to ensure that the best practice is followed.