Downloads · 30 days
16
16% of all-time downloads
nyasukun/qwen3-0.6b-cve-2025-3248-http-response-lora
qwen3-0.6b-cve-2025-3248-http-response-lora is a text generation model from nyasukun. Use it when you need the model to write or continue text. It is set up for peft.
LoRA adapter produced by the checked-in Agentic Model Trainer remote script.
Downloads · 30 days
16
16% of all-time downloads
All-time downloads
97
Public
Repo size
104 MB
Likes
0
Public
Click a slice to open those files.
.safetensors40.4 MB · 78%
From the Hugging Face model README
LoRA adapter produced by the checked-in Agentic Model Trainer remote script.
Base model: Qwen/Qwen3-0.6B@c1899de289a04d12100db370d81485cdf75e47ca
Datasets:
nyasukun/synthetic-CVE-2025-3248@dbcfb2186faaf5741818361747b2a65a10222c7fNormalized training rows: 3987
Normalized evaluation rows: 444
{
"excluded_input_fields": [
"method"
],
"input_fields": [
"path",
"body"
],
"input_schema": {
"body": "string",
"path": "string"
},
"objective": "causal_language_modeling",
"output_fields": [
"code",
"body"
],
"output_schema": {
"body": "string",
"code": "integer"
},
"pipeline_tag": "text-generation",
"prompt_renderer": "base_model_tokenizer_chat_template",
"prompt_renderer_options": {
"add_generation_prompt": true,
"enable_thinking": false
},
"request_grouping": "canonical_path_body_sha256",
"response_only_loss": true,
"response_termination": "eos_token",
"system_prompt": "Predict the HTTP response for the supplied request. Return only a JSON object with code and body fields.",
"wire_format": "json"
}
{
"conflict_rows_discarded": 552,
"conflict_variants_discarded": 550,
"conflicting_request_groups": 224,
"deduplication_strategy": "canonical_path_body_signature+one_response_per_request_mode_then_first_v2",
"duplicate_rows": 19,
"eval_response_code_counts": {
"200": 148,
"201": 9,
"202": 2,
"204": 4,
"302": 24,
"400": 50,
"401": 36,
"403": 105,
"404": 27,
"405": 1,
"422": 1,
"500": 37
},
"eval_unique_requests": 444,
"exact_duplicate_rows_removed": 19,
"group_split": true,
"http_conflict_policy": "mode_then_first",
"removed_rows": 569,
"retained_structured_rows": 4431,
"source_rows": 5000,
"train_response_code_counts": {
"200": 1331,
"201": 81,
"202": 22,
"204": 33,
"302": 213,
"400": 450,
"401": 329,
"403": 947,
"404": 243,
"405": 4,
"422": 4,
"500": 329,
"503": 1
},
"train_unique_requests": 3987,
"unique_request_response_pairs": 4981,
"unique_requests": 4431
}
{
"http_generation_metrics": {
"body_exact_match": 0.0,
"code_exact_match": 0.4,
"evaluated_requests": 5,
"joint_exact_match": 0.0,
"json_parse_rate": 1.0
},
"loss_metrics": {
"epoch": 3.0,
"eval_loss": 0.5778418779373169,
"eval_runtime": 6.0452,
"eval_samples_per_second": 73.447,
"eval_steps_per_second": 2.316
}
}
This adapter was produced with a Hugging Face Hub 1.x-compatible runtime. Use these mutually compatible pins in the Space's requirements.txt:
gradio==6.20.0
transformers==5.13.0
peft==0.19.1
huggingface-hub==1.5.0