TinyLlama-1.1B-Chat-v1.0
The TinyLlama project aims to pretrain a 1.1B Llama model on 3 trillion tokens. With some proper optimization, we can achieve this within a span of "just" 90 days using 16 A100-40G GPUs 🚀🚀. The training has started on 2023-09-01.
Params
1.1 B
Context
2,048
Downloads 30d
1.4 M
Likes
1,805
Download history
daily snapshots · 56 days
▲ 914 K in the last 30 days (38.9%)
2.3 M1.4 M
Aug 23Sep 2Sep 12Sep 21
2.6 M1.4 M
Jul 28Aug 15Sep 3Sep 21
Can you run it?
Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.
| File | Quant | Size | Est. VRAM | Verdict on RTX 4090 · 24 GB |
|---|---|---|---|---|
| model.safetensors | bf16 | 2.2 GB | 3.1 GB | ✅ Runs comfortably |
Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.
Run it
copy-paste, exact tags checked against the Hub$ curl -s https://aimodelscomparison.com/api/v1/models/tinyllama-1-1b-chat-v1-0
{
"hf_id": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
"params_b": 1.10,
"context_length": 2048,
"license": { "id": "apache-2.0", "commercial": "yes" },
"downloads_30d": 1438408,
"vram_estimates": [
{ "quant": "bf16", "gb": 3.1 }
],
"updated_at": "2026-07-28T17:07:40Z"
}
Specifications
- Architecture
- LlamaForCausalLM
- Parameters
- 1.1 B
- Tensor type
- BF16
- Context length
- 2,048
- Vocabulary
- 32,000
- Layers / heads
- 22 / 32
- Licence
- apache-2.0
- First seen on the Hub
- 2023-12-30
- Training datasets
- cerebras/SlimPajama-627B, bigcode/starcoderdata, HuggingFaceH4/ultrachat_200k, HuggingFaceH4/ultrafeedback_binarized
- Added to our catalog
- 2026-07-28
Compare with any text-generation model