What are the hardware requirements for Tora Code 13B V1.0?

Tora Code 13B V1.0 requires approximately 26 GB of VRAM and supports a context window of 16K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Tora Code 13B V1.0 and how large is it?

Tora Code 13B V1.0 is developed by llm-agents, a model with 13b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Tora Code 13B V1.0 perform on standard benchmarks?

Tora Code 13B V1.0 has the following published scores: MMLU 36.67. Compare against reference models on this page or on the LLM Explorer leaderboards.

Tora Code 13B V1.0 by llm-agents — VRAM 26GB, 16K context

Name: Tora Code 13B V1.0
Author: llm-agents

Tora Code 13B V1.0 is an open-source language model by llm-agents. Features: 13b LLM, VRAM: 26GB, Context: 16K, License: llama2, Code Generating, LLM Explorer Score: 0.13, Arc: 44.5, HellaSwag: 69.3, MMLU: 36.7, GSM8K: 8.2.

Arxiv:2309.17452 Code Codegen Dataset:competition math Dataset:gsm8k En Endpoints compatible Llama Math Pytorch Region:us Sharded

Model Card on HF 🤗: https://huggingface.co/llm-agents/tora-code-13b-v1.0

Tora Code 13B V1.0 Parameters and Internals

Model Type

text generation

Use Cases

Areas:

research, commercial applications

Applications:

mathematical reasoning, problem-solving

Primary Use Cases:

Mathematical problem solving using computation libraries and symbolic solvers

Considerations:

Combines natural language reasoning with computational tools

Additional Notes

ToRA integrates language reasoning and external computational tools for enhanced problem-solving capabilities

Supported Languages

en (fluent)

Training Details

Data Sources:

ToRA-Corpus 16k, MATH, GSM8k from GPT-4

Methodology:

Imitation learning (i.e., SFT) with output space shaping

Release Notes

Version:

v1.0

Date:

2023/10/08

Notes:

All ToRA models released at HuggingFace.

Version:

v1.0

Date:

2023/09/29

Notes:

ToRA paper, repo, and website released.

LLM Name	Tora Code 13B V1.0
Repository 🤗	https://huggingface.co/llm-agents/tora-code-13b-v1.0
Model Size	13b
Required VRAM	26 GB
Updated	2026-06-02
Maintainer	llm-agents
Model Type	llama
Model Files	9.9 GB: 1-of-3 9.9 GB: 2-of-3 6.2 GB: 3-of-3
Supported Languages	en
Generates Code	Yes
Model Architecture	LlamaForCausalLM
License	llama2
Context Length	16384
Model Max Length	16384
Transformers Version	4.31.0
Tokenizer Class	LlamaTokenizer
Beginning of Sentence Token	<s>
End of Sentence Token	</s>
Unk Token	<unk>
Vocabulary Size	32001
Torch Data Type	bfloat16

Rank the Tora Code 13B V1.0 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

Quantized Models of the Tora Code 13B V1.0

Model	Likes	Downloads	VRAM
Tora Code 13B V1.0 GGUF	4	370	5 GB
Tora Code 13B V1.0 AWQ	1	11	7 GB
Tora Code 13B V1.0 GPTQ	0	7	7 GB

Best Alternatives to Tora Code 13B V1.0

Best Alternatives	Context / RAM	Downloads	Likes
NexusRaven V2 13B	16K / 26 GB	105	472
CodeLlama 13B Instruct Hf	16K / 26 GB	2469	160
CodeLlama 13B MORepair	16K / 26 GB	9	2
CodeLlama 13B Hf	16K / 26 GB	3466	117
CodeLlama 13B Python Hf	16K / 26 GB	558	51
...ma 13B Hf Truncated Embeddings	16K / 52.3 GB	5	0
...ma Airoboros Orca Platypus 13B	16K / 26 GB	117	0
Codellama 13B Oasst Sft V10	16K / 25.9 GB	184	70
CodeLlama 13B Hf	16K / 26 GB	224	0
CodeLlama 13B Instruct Hf	16K / 26 GB	369	29

Note: green Score (e.g. "73.2") means that the model is better than llm-agents/tora-code-13b-v1.0.