ld_requests / deepset /roberta-large-squad2_quant_request_False_W4A16_4bit_int4.json
n1ck-guo's picture
Add deepset/roberta-large-squad2 to quant queue
25d5784 verified
raw
history blame contribute delete
926 Bytes
{"job_type": "quantization & evaluation", "model": "deepset/roberta-large-squad2", "revision": "main", "private": false, "params": 0.35, "architectures": "RobertaForQuestionAnswering", "num_layers": 24, "input_dtype": "float16", "input_bits": 16, "model_params": 0.35, "model_weight_gb": 1.42, "quant_scheme": "INT4 (W4A16)", "quant_precision": "4bit", "quant_weight_dtype": "int4", "quant_bits": 4, "quant_memory_gb": 0.15, "quant_gpu_type": "NVIDIA GeForce RTX 4090", "quant_gpu_nums": 1, "eval_memory_gb": 0.21, "eval_gpu_type": "NVIDIA GeForce RTX 4090", "eval_gpu_nums": 1, "quant_model_size_gb": 0.17, "hardware": "NVIDIA GeForce RTX 4090", "gpu_nums": 1, "script": "auto_quant", "status": "Pending", "submitted_time": "2026-04-29T13:01:35Z", "model_type": "quantization", "job_id": -1, "job_start_time": null, "likes": 29, "license": "cc-by-4.0", "still_on_hub": true, "tags": ["moe"], "downloads": 0, "created_at": ""}