
Base model	`unsloth/Qwen3.5-9B`
Architecture	`Qwen3_5ForConditionalGeneration`
Parameters	9.7B
Training data	`ermiaazarkhalili/Fable-5-Glint-Clean` (private)
Method	LoRA supervised fine-tuning via Unsloth + TRL
License	`apache-2.0` (inherited from the base model)

Usage

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "ermiaazarkhalili/Qwen3.5-9B-SFT-Fable5-Glint"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, dtype='auto', device_map='auto')

messages = [{"role": "user", "content": "Explain gradient checkpointing in two sentences."}]
inputs = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True, return_tensors='pt'
).to(model.device)

outputs = model.generate(inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training configuration

Setting	Value
LoRA rank (r)	16
LoRA alpha	16
Learning rate	0.0002
Epochs	3
Effective batch size	8 (1 x 8 grad accum)
Max sequence length	4096
Base precision	4-bit (QLoRA)
Target modules	`q_proj`, `k_proj`, `v_proj`, `o_proj`, `gate_proj`, `up_proj`, `down_proj`, `out_proj`

Held-out evaluation

Next-token accuracy on a deterministic held-out split of ermiaazarkhalili/Fable-5-Glint-Clean (n = 199 samples), scored against the base model unsloth/Qwen3.5-9B. Assistant tokens only; both models are scored identically.

Metric	Base	This model	Δ
Top-1 accuracy	0.6461	0.7329	+0.0869
Top-5 accuracy	0.8842	0.9385	+0.0543

A delta measures how far fine-tuning moved this model from its own starting point; it is not a ranking against other models, which start from different baselines.

Observed training loss

Measured from our SLURM logs for this configuration. These are training-loss observations only — see the held-out evaluation above for measured accuracy.

SLURM job	Steps	First loss	Final loss
`46021013`	1,554	0.9945	0.6509
`45987993`	300	1.1245	0.8734

Limitations

No benchmark evaluation has been run on this checkpoint. The only reported numbers are training-loss observations.
Inherits the biases, knowledge cutoff and failure modes of the base model.
Fine-tuned on a single instruction-following dataset; behaviour outside that distribution is untested.
LoRA adapters were merged into the base weights, so the merged model cannot be detached from this fine-tune.

Reproducing

Trained by notebooks/fable_distillation_qwen35-9b_fable-glint_unsloth.ipynb, executed non-interactively with papermill on a SLURM H100 partition (Unsloth + TRL, LoRA).

Card generated from the training run's own configuration and logs by scripts/generate_hub_model_card.py.