Text Generation
PEFT
Safetensors
English
code
cobol
code-generation
mainframe-modernization
lora
sft
trl
unsloth
ministral
conversational
Eval Results (legacy)
Instructions to use axeltta/mistral-axel-1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use axeltta/mistral-axel-1 with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("unsloth/ministral-3-8b-instruct-2512-unsloth-bnb-4bit") model = PeftModel.from_pretrained(base_model, "axeltta/mistral-axel-1") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Unsloth Desktop
Invalid JSON:Unexpected token 'N', ..."al_loss": NaN,
""... is not valid JSON
| { | |
| "model": "unsloth/Ministral-3-8B-Instruct-2512", | |
| "lora_r": 16, | |
| "lora_alpha": 16, | |
| "max_steps": 40, | |
| "learning_rate": 0.0001, | |
| "batch_size": 4, | |
| "seed": 12627998, | |
| "train_n": 95, | |
| "val_n": 10, | |
| "eval_metrics": { | |
| "eval_loss": NaN, | |
| "eval_runtime": 5.3084, | |
| "eval_samples_per_second": 1.884, | |
| "eval_steps_per_second": 0.565, | |
| "epoch": 1.6666666666666665 | |
| }, | |
| "train_runtime_s": 65.463, | |
| "adapter_path": "/adapters/cobol-string-sft" | |
| } |