What are the hardware requirements for Codegen 350M Mono?

Codegen 350M Mono requires approximately 0.8 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Codegen 350M Mono and how large is it?

Codegen 350M Mono is developed by Salesforce, a model with 350m 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 Codegen 350M Mono?

Codegen 350M Mono is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Codegen 350M Mono by Salesforce — VRAM 0.8GB

Name: Codegen 350M Mono
Author: Salesforce

Codegen 350M Mono is an open-source language model by Salesforce. Features: 350m LLM, VRAM: 0.8GB, License: bsd-3-clause, Code Generating, LLM Explorer Score: 0.07.

Arxiv:2203.13474 Codegen Deploy:azure Endpoints compatible Pytorch Region:us

Model Card on HF 🤗: https://huggingface.co/Salesforce/codegen-350M-mono

Codegen 350M Mono Parameters and Internals

Model Type

autoregressive language model, code generation

Use Cases

Areas:

program synthesis

Applications:

code completion, feature extraction from programming language texts

Primary Use Cases:

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

Supported Languages

python (Native)

Training Details

Data Sources:

BigPython dataset

Data Volume:

71.7B tokens

Methodology:

cross-entropy loss to maximize the likelihood of sequential inputs

Hardware Used:

TPU-v4-512

Model Architecture:

autoregressive

Input Output

Input Format:

comment string

Accepted Modalities:

text

Output Format:

executable code

LLM Name	Codegen 350M Mono
Repository 🤗	https://huggingface.co/Salesforce/codegen-350M-mono
Model Size	350m
Required VRAM	0.8 GB
Updated	2026-07-16
Maintainer	Salesforce
Model Type	codegen
Model Files	0.8 GB
Generates Code	Yes
Model Architecture	CodeGenForCausalLM
License	bsd-3-clause
Model Max Length	2048
Transformers Version	4.21.0.dev0
Tokenizer Class	GPT2Tokenizer
Vocabulary Size	51200
Torch Data Type	float16
Activation Function	gelu_new

Rank the Codegen 350M Mono 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 Codegen 350M Mono

Best Alternatives	Context / RAM	Downloads	Likes
Ft Calculator	0K / 0.6 GB	5	0
...o Python 18K Alpaca Runpod.net	0K / 0.7 GB	2	1
...hs Finetuned Python 18K Alpaca	0K / 0.8 GB	5	0
...no Finetuned Python 18K Alpaca	0K / 0.8 GB	6	1
... Powershell Codegen 350M Multi	0K / 1.4 GB	4	2
Codegen 350M Mono Java Merged	0K / 1.5 GB	11	0
...en 350M Mono 18K Alpaca Python	0K / 0.7 GB	8	2
Nsql 350M	0K / 1.5 GB	63	33
Codegen 350M Mono	0K / GB	488	6
Codegen 350M Multi	0K / GB	182	2

Note: green Score (e.g. "73.2") means that the model is better than Salesforce/codegen-350M-mono.