LLM EXPLORER 65,444 MODELS INDEXED

Phind CodeLlama 34B Python V1 GPTQ by TheBloke

By TheBloke · 16 downloads
HF Repo ↗ Copy URL

Phind CodeLlama 34B Python V1 GPTQ is an open-source language model by TheBloke. Features: 34b LLM, VRAM: 17.7GB, Context: 16K, License: llama2, Quantized, Code Generating, LLM Explorer Score: 0.08.

  4-bit Base model:phind/phind-codella... Base model:quantized:phind/phi...   Code llama   Codegen   Gptq   Llama   Model-index   Quantized   Region:us   Safetensors
Model Card on HF 🤗: https://huggingface.co/TheBloke/Phind-CodeLlama-34B-Python-v1-GPTQ 

Phind CodeLlama 34B Python V1 GPTQ Parameters and Internals

Model Type 
text-generation, Llama
Use Cases 
Areas:
Research, Programming
Applications:
Code completion, Automated code generation
Primary Use Cases:
HumanEval tasks
Limitations:
Limited testing, Additional safety testing required
Additional Notes 
The model is provided in multiple quantized formats for different hardware and efficiency requirements. Act-order and group-size options are included to optimize for VRAM usage and performance accuracy.
Supported Languages 
Python (High proficiency)
Training Details 
Data Sources:
Proprietary dataset of ~80k high quality programming problems and solutions
Data Volume:
~160k examples shown
Methodology:
Fine-tuned from CodeLlama-34B-Python for 2 epochs, No LoRA, Trained using DeepSpeed ZeRO 3 and Flash Attention 2
Context Length:
4096
Training Time:
3 hours
Hardware Used:
32x A100-80GB GPUs
Model Architecture:
Fine-tuned on internal dataset to achieve high pass rate on HumanEval
Safety Evaluation 
Methodologies:
Limited testing
Input Output 
Input Format:
Plain text prompt with '\n' at the end of your task
Accepted Modalities:
text
Output Format:
Generated text
LLM NamePhind CodeLlama 34B Python V1 GPTQ
Repository 🤗https://huggingface.co/TheBloke/Phind-CodeLlama-34B-Python-v1-GPTQ 
Model CreatorPhind
Base Model(s)  Phind CodeLlama 34B Python V1   Phind/Phind-CodeLlama-34B-Python-v1
Model Size34b
Required VRAM17.7 GB
Updated2026-08-03
MaintainerTheBloke
Model Typellama
Model Files  17.7 GB
GPTQ QuantizationYes
Quantization Typegptq
Generates CodeYes
Model ArchitectureLlamaForCausalLM
Licensellama2
Context Length16384
Model Max Length16384
Transformers Version4.33.0.dev0
Tokenizer ClassLlamaTokenizer
Beginning of Sentence Token<s>
End of Sentence Token</s>
Unk Token<unk>
Vocabulary Size32000
Torch Data Typebfloat16

Best Alternatives to Phind CodeLlama 34B Python V1 GPTQ

Best Alternatives
Context / RAM
Downloads
Likes
...chless Codellama 34B V2.0 GPTQ16K / 17.7 GB64
CodeFuse CodeLlama 34B 4bits16K / 19 GB2626
...nbuddy Coder 34B V11 Bf16 GPTQ16K / 17.9 GB90
Phind CodeLlama 34B V2 GPTQ16K / 17.7 GB6490
CodeLlama 34B Instruct GPTQ16K / 18.3 GB15074
MAmmoTH Coder 34B GPTQ16K / 17.7 GB102
Synthia 34B V1.2 GPTQ16K / 17.7 GB97
CodeFuse CodeLlama 34B GPTQ16K / 17.7 GB49
...zardCoder Python 34B V1.0 GPTQ16K / 17.7 GB5562
CodeLlama 34B Guanaco Gptq16K / 18.3 GB655
Note: green Score (e.g. "73.2") means that the model is better than TheBloke/Phind-CodeLlama-34B-Python-v1-GPTQ.