rostlabs/rost-1b-base

Update README.md

Add English evaluation battery (lm-evaluation-harness 0.4.12, zero-shot + 5-shot MMLU)

The weights are bfloat16 now; say so instead of explaining float32

Remove the float32 shards, superseded by model.safetensors

Remove the float32 shards, superseded by model.safetensors

Remove the float32 shards, superseded by model.safetensors

Declare bfloat16

Publish the weights in bfloat16, the precision they were trained in

Drop the last link into a private repository

The fork moved to the rostlabs org; drop links into a private repo

GGUF and llama.cpp are available now; say which runtimes are not

Say which decoding settings to use, and why 1.1

Ship the decoding settings, and a stop condition

Rewrite the card: evaluation, intended use, limitations

Update README.md

Upload tokenizer/tokenizer.pkl with huggingface_hub

Upload tokenizer/token_bytes.pt with huggingface_hub

Upload folder using huggingface_hub

Upload meta_011136.json with huggingface_hub

Upload model_011136.pt with huggingface_hub

Upload folder using huggingface_hub

initial commit