Update README.md
Browse files
README.md
CHANGED
@@ -9,4 +9,69 @@ library_name: transformers
|
|
9 |
tags:
|
10 |
- coder
|
11 |
- text-generation-inference
|
12 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
9 |
tags:
|
10 |
- coder
|
11 |
- text-generation-inference
|
12 |
+
---
|
13 |
+
![coderx.png](https://cdn-uploads.huggingface.co/production/uploads/65bb837dbfb878f46c77de4c/ncJZH_SSIpEr16oAq4qDF.png)
|
14 |
+
# **Viper-Coder-v0.1-GGUF**
|
15 |
+
|
16 |
+
Viper-Coder-v0.1-GGUF is based on the Qwen 2.5 14B modality architecture, designed to enhance the reasoning capabilities of 14B-parameter models. It has been fine-tuned on a synthetic dataset based on the latest coding logits and CoT datasets, further optimizing its chain-of-thought (CoT) reasoning and logical problem-solving abilities. The model demonstrates significant improvements in context understanding, structured data processing, and long-context comprehension, making it ideal for complex reasoning tasks, instruction-following, and text generation.
|
17 |
+
|
18 |
+
### **Key Improvements**
|
19 |
+
1. **Enhanced Knowledge and Expertise**: Improved mathematical reasoning, coding proficiency, and structured data processing.
|
20 |
+
2. **Fine-Tuned Instruction Following**: Optimized for precise responses, structured outputs (e.g., JSON), and generating long texts (8K+ tokens).
|
21 |
+
3. **Greater Adaptability**: Better role-playing capabilities and resilience to diverse system prompts.
|
22 |
+
4. **Long-Context Support**: Handles up to **128K tokens** and generates up to **8K tokens** per output.
|
23 |
+
5. **Multilingual Proficiency**: Supports over **29 languages**, including Chinese, English, French, Spanish, Portuguese, German, and more.
|
24 |
+
|
25 |
+
### **Quickstart with Transformers**
|
26 |
+
|
27 |
+
```python
|
28 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
29 |
+
|
30 |
+
model_name = "prithivMLmods/Viper-Coder-v0.1-GGUF"
|
31 |
+
|
32 |
+
model = AutoModelForCausalLM.from_pretrained(
|
33 |
+
model_name,
|
34 |
+
torch_dtype="auto",
|
35 |
+
device_map="auto",
|
36 |
+
trust_remote_code=True
|
37 |
+
)
|
38 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
39 |
+
|
40 |
+
prompt = "Give me a short introduction to large language models."
|
41 |
+
messages = [
|
42 |
+
{"role": "system", "content": "You are an advanced AI assistant with expert-level reasoning and knowledge."},
|
43 |
+
{"role": "user", "content": prompt}
|
44 |
+
]
|
45 |
+
text = tokenizer.apply_chat_template(
|
46 |
+
messages,
|
47 |
+
tokenize=False,
|
48 |
+
add_generation_prompt=True
|
49 |
+
)
|
50 |
+
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
|
51 |
+
|
52 |
+
generated_ids = model.generate(
|
53 |
+
**model_inputs,
|
54 |
+
max_new_tokens=512
|
55 |
+
)
|
56 |
+
generated_ids = [
|
57 |
+
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
|
58 |
+
]
|
59 |
+
|
60 |
+
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
61 |
+
print(response)
|
62 |
+
```
|
63 |
+
|
64 |
+
### **Intended Use**
|
65 |
+
- **Advanced Reasoning & Context Understanding**: Designed for logical deduction, multi-step problem-solving, and complex knowledge-based tasks.
|
66 |
+
- **Mathematical & Scientific Problem-Solving**: Enhanced capabilities for calculations, theorem proving, and scientific queries.
|
67 |
+
- **Code Generation & Debugging**: Generates and optimizes code across multiple programming languages.
|
68 |
+
- **Structured Data Analysis**: Processes tables, JSON, and structured outputs, making it ideal for data-centric tasks.
|
69 |
+
- **Multilingual Applications**: High proficiency in over 29 languages, enabling global-scale applications.
|
70 |
+
- **Extended Content Generation**: Supports detailed document writing, research reports, and instructional guides.
|
71 |
+
|
72 |
+
### **Limitations**
|
73 |
+
1. **High Computational Requirements**: Due to its **14B parameters** and **128K context support**, it requires powerful GPUs or TPUs for efficient inference.
|
74 |
+
2. **Language-Specific Variability**: Performance may vary across supported languages, especially for low-resource languages.
|
75 |
+
3. **Potential Error Accumulation**: Long-text generation can sometimes introduce inconsistencies over extended outputs.
|
76 |
+
4. **Limited Real-World Awareness**: Knowledge is restricted to training data and may not reflect recent world events.
|
77 |
+
5. **Prompt Sensitivity**: Outputs can depend on the specificity and clarity of the input prompt.
|