Update README.md
6e0b3d4 verified - chkpt Format-blended SFT: chat fluency + preserved benchmark formats
- 1.52 kB initial commit
- 15 kB Update README.md
- 4.82 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 5.91 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 22 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 3.34 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 19 MB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 15.5 kB generate.py: nrng decoding defaults (temp 0.7, rep-pen 1.15) -- best per decode_lab_chat
- 11.6 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 3.1 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 40 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 9.7 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 416 Bytes Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 3.17 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 5.71 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 263 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 14 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 10.7 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
- 17.9 kB Add Mycel-LM 79M: architecture, tokenizer, training scripts, weights-only checkpoints
