What are the hardware requirements for Codegen2 3 7B P?

Codegen2 3 7B P requires approximately 14.6 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Codegen2 3 7B P and how large is it?

Codegen2 3 7B P is developed by Salesforce, a model with 7b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Codegen2 3 7B P?

Codegen2 3 7B P is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Codegen2 3 7B P by Salesforce — VRAM 14.6GB

Name: Codegen2 3 7B P
Author: Salesforce

Codegen2 3 7B P is an open-source language model by Salesforce. Features: 7b LLM, VRAM: 14.6GB, License: apache-2.0, Code Generating, LLM Explorer Score: 0.07.

Arxiv:2305.02309 Codegen Custom code Endpoints compatible Pytorch Region:us Sharded

Model Card on HF 🤗: https://huggingface.co/Salesforce/codegen2-3_7B_P

Codegen2 3 7B P Parameters and Internals

Model Type

program synthesis, infilling

Use Cases

Areas:

program synthesis

Applications:

program synthesis

Primary Use Cases:

generating executable code given English prompts, completing partially-generated code

Limitations:

program synthesis limitation

Supported Languages

c (unknown), c++ (unknown), c-sharp (unknown), dart (unknown), go (unknown), java (unknown), javascript (unknown), kotlin (unknown), lua (unknown), php (unknown), python (unknown), ruby (unknown), rust (unknown), scala (unknown), shell (unknown), sql (unknown), swift (unknown), typescript (unknown), vue (unknown)

Training Details

Data Sources:

stricter permissive subset of the deduplicated version of the Stack dataset (v1.1)

Methodology:

cross-entropy loss for causal language modeling and file-level span corruption

LLM Name	Codegen2 3 7B P
Repository 🤗	https://huggingface.co/Salesforce/codegen2-3_7B_P
Model Size	7b
Required VRAM	14.6 GB
Updated	2026-08-06
Maintainer	Salesforce
Model Type	codegen
Model Files	9.9 GB: 1-of-2 4.7 GB: 2-of-2
Generates Code	Yes
Model Architecture	CodeGenForCausalLM
License	apache-2.0
Model Max Length	1024
Transformers Version	4.25.1
Tokenizer Class	GPT2Tokenizer
Vocabulary Size	51200
Torch Data Type	float32
Activation Function	gelu_new

Rank the Codegen2 3 7B P capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Codegen2 3 7B P

Best Alternatives	Context / RAM	Downloads	Likes
Codegen2 7B P	0K / 27.5 GB	93	26

Note: green Score (e.g. "73.2") means that the model is better than Salesforce/codegen2-3_7B_P.