samai-9b is an open-weight model from Tchbcn. Its published files total 32.8 GB.
Parameters—
Context—
Weights32.5 GB
License—
AccessOpen weights
Monthly Downloads—
Model Card
The publisher has not written a card for this model.
Identity and Version
- Repository
- tchbcb/samai-9b
- Publisher
- Tchbcn
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- bef41ea267f773ec6d19e9109c9e07349a2168d4
- First published
- 2026-09-22
- Last updated
- 2026-09-23
Files and Weights
33 files, 32.8 GB in total. The weights are 5 files totalling 32.5 GB in gguf, safetensors.
Weights5 files · 32.5 GB
Configuration4 files · 5.3 KB
Tokenizer4 files · 40.0 MB
Documentation4 files · 42.1 KB
Other15 files · 203.7 MB
Repository1 file · 1.8 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| m13_r8_lora/adapter_model.safetensors | Weights | 232.8 MB | 455f0ebdcba2 |
| m13_r8_mixbit_4x8.gguf | Weights | 6.9 GB | 16de0ceee042 |
| m13_r8_q4_k_m.gguf | Weights | 5.6 GB | 196ba3e75fdb |
| merged_m13_r8_hf/model.safetensors | Weights | 18.8 GB | 725d69685d2c |
| mmproj_m11.gguf | Weights | 921.7 MB | d89c4bc142d0 |
| m13_r8_lora/adapter_config.json | Configuration | 1.1 KB | — |
| merged_m13_r8_hf/config.json | Configuration | 2.9 KB | — |
| merged_m13_r8_hf/generation_config.json | Configuration | 116 B | — |
| merged_m13_r8_hf/processor_config.json | Configuration | 1.2 KB | — |
| README.md | Documentation | — | |
| artifacts/m15_reports/R10_verdict.md | Documentation | 3.8 KB | — |
| m13_r8_lora/README.md | Documentation | 229 B | — |
| train.md | Documentation | 38.0 KB | — |
| artifacts/m15_reports/MIXBIT_BUILD_DONE | Other | 290 B | — |
| artifacts/m15_reports/mixbit_build.log | Other | 7.4 KB | — |
| artifacts/m15_reports/probe_long_f16long.txt | Other | 955 B | — |
| artifacts/m15_reports/probe_long_mixbit4x8long.txt | Other | 970 B | — |
| artifacts/m15_reports/probe_long_mixbithighlong.txt | Other | 1.0 KB | — |
| artifacts/m15_reports/probe_serving_mixbit4x8.txt | Other | 1.7 KB | — |
| artifacts/m15_reports/probe_serving_mixbit4x8b2.txt | Other | 1.7 KB | — |
| artifacts/m15_reports/probe_serving_mixbithigh.txt | Other | 2.2 KB | — |
| artifacts/m15_reports/probe_serving_q4_baseline.txt | Other | 2.0 KB | — |
| llama_cpp_cuda_t4_v2.tar.gz | Other | 144.2 MB | f526dc4e05b7 |
| m13_r8_lora/chat_template.jinja | Other | 7.8 KB | — |
| m13_r8_q4_sha256.txt | Other | 65 B | — |
| merged_m13_r8_hf/chat_template.jinja | Other | 7.8 KB | — |
| zagent_fixed.tar.gz | Other | 29.9 MB | d21b7b9c4ed3 |
| zagent_r11.tar.gz | Other | 29.6 MB | 9772636dccda |
| .gitattributes | Repository | 1.8 KB | — |
| m13_r8_lora/tokenizer.json | Tokenizer | 20.0 MB | 06b9509352d2 |
| m13_r8_lora/tokenizer_config.json | Tokenizer | 1.1 KB | — |
| merged_m13_r8_hf/tokenizer.json | Tokenizer | 20.0 MB | 06b9509352d2 |
| merged_m13_r8_hf/tokenizer_config.json | Tokenizer | 1.2 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 32.5 GB
Download from Tchbcn
Released by Tchbcn through its official repository on Hugging Face.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 32.5 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.