Downloads · 30 days
8
20% of all-time downloads
charlie-li/Qwen3-30B-A3B-Instruct-2507-ScaleSWE-Distilled-Epoch3-v2
Qwen3-30B-A3B-Instruct-2507-ScaleSWE-Distilled-Epoch3-v2 is a machine learning model from charlie-li. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
This directory contains the Epoch 3 v2 ScaleSWE-Distilled SFT checkpoint.
Downloads · 30 days
8
20% of all-time downloads
All-time downloads
40
Public
Parameters
30.5B
61.1 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors61.1 GB · 100%
From the Hugging Face model README
This directory contains the Epoch 3 v2 ScaleSWE-Distilled SFT checkpoint.
This v2 release is a continued full-parameter fine-tune that starts from the prior
charlie-li/Qwen3-30B-A3B-Instruct-2507-ScaleSWE-Distilled-Epoch3 package and adapts the model to the longer-context
(70k → 128k tokens) ScaleSWE-Distilled trajectories. The original pretrained base model is
Qwen/Qwen3-30B-A3B-Instruct-2507.
charlie-li/Qwen3-30B-A3B-Instruct-2507-ScaleSWE-Distilled-Epoch3Qwen3-30B-A3B-Instruct-2507Qwen3MoeForCausalLMdataset_index.yamltraining_config.yamlscaleswe_distilled_sharegpt_format_nothink_tool_role_70k_to_128k.v1_messages.jsonl4mp_replicate_size=4, mp_shard_size=8131072 tokens11281e-50.05bfloat16Qwen/Qwen3-30B-A3B-Instruct-2507/30 through 2 (zero-based epoch counter)50 steps5207 / 2070.3053124248981476model-00001-of-00016.safetensors through model-00016-of-00016.safetensorsmodel.safetensors.index.jsonconfig.jsongeneration_config.jsontokenizer.json, tokenizer_config.jsonchat_template.jinjatrainer_log.jsonltraining_config.yaml, dataset_index.yamlfrom transformers import AutoModelForCausalLM, AutoTokenizer
model_dir = "charlie-li/Qwen3-30B-A3B-Instruct-2507-ScaleSWE-Distilled-Epoch3-v2"
tokenizer = AutoTokenizer.from_pretrained(model_dir, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
model_dir,
trust_remote_code=True,
device_map="auto",
)