Feature Extraction
Transformers
Safetensors
autojev_decision
jev
decision
autojev
typed-decision
classification
distillation
custom_code
Instructions to use inclusionAI/LemonDecider with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use inclusionAI/LemonDecider with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="inclusionAI/LemonDecider", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("inclusionAI/LemonDecider", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Sanitize metadata: remove internal experiment paths and codenames
Browse files- config.json +2 -2
- sft_receipt.json +1 -2
config.json
CHANGED
|
@@ -270,9 +270,9 @@
|
|
| 270 |
"initialization_step": 24,
|
| 271 |
"max_input_length": 8192,
|
| 272 |
"max_options": 255,
|
| 273 |
-
"model_id": "
|
| 274 |
"model_type": "autojev_decision",
|
| 275 |
-
"project_name": "
|
| 276 |
"temperature": 1.0,
|
| 277 |
"text_config": {
|
| 278 |
"attention_bias": false,
|
|
|
|
| 270 |
"initialization_step": 24,
|
| 271 |
"max_input_length": 8192,
|
| 272 |
"max_options": 255,
|
| 273 |
+
"model_id": "lemondecider-qwen3.5-4b-v1",
|
| 274 |
"model_type": "autojev_decision",
|
| 275 |
+
"project_name": "lemondecider",
|
| 276 |
"temperature": 1.0,
|
| 277 |
"text_config": {
|
| 278 |
"attention_bias": false,
|
sft_receipt.json
CHANGED
|
@@ -5,7 +5,6 @@
|
|
| 5 |
"best_dev_accuracy_t1": 0.7804877758026123,
|
| 6 |
"lr": 2.2827515498423204e-07,
|
| 7 |
"examples_seen": 6144,
|
| 8 |
-
"source_checkpoint": "/mnt/secbasemodel-p/common/dev/543950/lemonjev/distill-v1/runs/lemonjev-imajev4b-distill-goldfix-v1-nativeprompt/checkpoint-00101",
|
| 9 |
"temperature_fit": {
|
| 10 |
"method": "NLL scan on dev 574 under fork _known() semantics, T grid [0.30,5.00]x0.05 (from cached T=1 probs)",
|
| 11 |
"best_T": 0.6,
|
|
@@ -14,6 +13,6 @@
|
|
| 14 |
"dev_nll_at_1": 0.6016,
|
| 15 |
"dev_brier_at_1": 0.29885,
|
| 16 |
"bound": false,
|
| 17 |
-
"reason": "best_T=0.60 != serving T=1.0; NOT applied —
|
| 18 |
}
|
| 19 |
}
|
|
|
|
| 5 |
"best_dev_accuracy_t1": 0.7804877758026123,
|
| 6 |
"lr": 2.2827515498423204e-07,
|
| 7 |
"examples_seen": 6144,
|
|
|
|
| 8 |
"temperature_fit": {
|
| 9 |
"method": "NLL scan on dev 574 under fork _known() semantics, T grid [0.30,5.00]x0.05 (from cached T=1 probs)",
|
| 10 |
"best_T": 0.6,
|
|
|
|
| 13 |
"dev_nll_at_1": 0.6016,
|
| 14 |
"dev_brier_at_1": 0.29885,
|
| 15 |
"bound": false,
|
| 16 |
+
"reason": "best_T=0.60 != serving T=1.0; NOT applied — public-set re-eval required before adoption (dev acc unchanged at T*, argmax-invariant except noul boundary)"
|
| 17 |
}
|
| 18 |
}
|