---
license: apache-2.0
pipeline_tag: text-to-speech
library_name: rlx
tags:
- sesame
- csm
- tts
- rlx
---

# Sesame CSM-1B (RLX staging)

CSM-1B conversational TTS weights (ungated mirror) for RLX.

| Field | Value |
|---|---|
| **Hub id** | [`eugenehp/sesame`](https://huggingface.co/eugenehp/sesame) |
| **Kind** | Staging redistrib of an upstream checkpoint for RLX runners. |
| **RLX crate** | [`rlx-sesame`](https://github.com/MIT-RLX/rlx-models/tree/main/crates/rlx-sesame) |
| **Upstream** | https://huggingface.co/unsloth/csm-1b |

## Quick start

```bash
hf download eugenehp/sesame --local-dir .
cargo run -p rlx-sesame --release -- --model-dir .
```

## File highlights

- `model.safetensors` (3.9 GiB)
- `sesame-csm-backbone.gguf` (1.2 GiB)
- `tokenizer.json` (16.4 MiB)
- `tokenizer_config.json` (49.4 KiB)
- `config.json` (3.0 KiB)
- `chat_template.jinja` (2.0 KiB)
- `special_tokens_map.json` (449 B)
- `preprocessor_config.json` (271 B)
- `generation_config.json` (264 B)

## Run with RLX

Clone [rlx-models](https://github.com/MIT-RLX/rlx-models), place this repo under `weights/tts/sesame` (or pass the path explicitly), then:

```bash
cargo run -p rlx-sesame --release -- --model-dir .
```

## License

Apache License 2.0 — see `LICENSE`. Inherit upstream terms when redistributing.

Original weights and authorship: https://huggingface.co/unsloth/csm-1b

## Redistrib note

This Hub repo exists so RLX recipes have a stable fetch target. When you only need the upstream checkpoint, prefer the Upstream link above.