Skip to page content
SAVRN
Search Plan a project

Open-weight model

AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b

by Hong Ting Tsang gzone0111/AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b

AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b is an open-weight model from Hong Ting Tsang. It has 7.6B parameters and a 32,768-token context. At 16-bit it needs about 18.3 GB of GPU memory, which fits on 1x MI300X from $1.85 an hour, at the lowest prices in the SAVRN Index.

Parameters7.6B
Context32,768
Weights15.2 GB
License—
AccessOpen weights
Monthly Downloads—

Runs On

What it takes to serve AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b (7.6B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 15.2 GB 18.3 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 7.6 GB 9.1 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 3.8 GB 4.6 GB 1x MI300X (192 GB)
Vultr
$1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Oct 18, 2026.

AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b on every accelerator the SAVRN Index prices, at every precision

Model Card

The publisher has not written a card for this model.

Configuration

Architecture
Qwen2ForCausalLM
Context length (tokens)
32,768
Layers
28
Hidden size
3,584
Feed-forward size
18,944
Attention heads
28
Key/value heads
4
Vocabulary size
152,064
RoPE base
1e+06
Model type
qwen2

Identity and Version

Repository
gzone0111/AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b
Publisher
Hong Ting Tsang
Task
Not stated by the source
Modality
Other
Library
Not stated by the source
Parameters
7.6B parameters
Languages
Not stated by the source
Revision
5ada71b409e6f42af032cff9a8b02b2101248cab
First published
2026-09-24
Last updated
2026-09-25

Files and Weights

112 files, 22.5 GB in total. The weights are 4 files totalling 15.2 GB in safetensors.

Weights4 files · 15.2 GB
Configuration24 files · 137.6 MB
Tokenizer4 files · 15.9 MB
Other79 files · 7.1 GB
Repository1 file · 5.1 KB
Every file
FileTypeSizeSHA-256
model-00001-of-00004.safetensorsWeights4.7 GB 6199056e7cc4
model-00002-of-00004.safetensorsWeights5.0 GB 3cb5c6649817
model-00003-of-00004.safetensorsWeights5.0 GB 21277e81eee2
model-00004-of-00004.safetensorsWeights572.5 MB bf2efdf9ba87
added_tokens.jsonConfiguration605 B —
benchmark/graph_retrieval/2wikimultihopqa/result_20251121170019_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration5.0 MB —
benchmark/graph_retrieval/2wikimultihopqa/summary_20251121170019_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration539 B —
benchmark/graph_retrieval/hotpotqa/result_20251122003220_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration5.2 MB —
benchmark/graph_retrieval/hotpotqa/summary_20251122003220_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration567 B —
benchmark/graph_retrieval/musique/result_20251121205919_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration7.6 MB —
benchmark/graph_retrieval/musique/summary_20251121205919_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration568 B —
benchmark/graph_retrieval/nq/result_20251122034015_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration4.4 MB —
benchmark/graph_retrieval/nq/summary_20251122034015_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration535 B —
benchmark/graph_retrieval/popqa/result_20251122055250_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration4.1 MB —
benchmark/graph_retrieval/popqa/summary_20251122055250_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_Qwen2.5-7B-Instruct.jsonConfiguration536 B —
config.jsonConfiguration1.3 KB —
constructed_kg/2021wiki_output/kg_extraction/_data_httsangaj_autograph_checkpoints_ablation_judge_qwen7b-1.5b-judge_global_step_50_actor_huggingface_2021wiki_output_20251121084048_1_in_1.jsonConfiguration25.7 MB a8319be4ca59
constructed_kg/2wikimultihopqa_output/kg_extraction/_root_data_checkpoints_20251120_123621_qwen2.5-7b-judge-1.5B-autograph-distract_easy-docsize15-textlinkingFalse-f1False_global_step_50_actor_huggingface_2wikimultihopqa_output_20251120184858_1_in_1.jsonConfiguration16.2 MB a8e56b3b71b9
constructed_kg/hotpotqa_output/kg_extraction/_data_httsangaj_autograph_checkpoints_ablation_judge_qwen7b-1.5b-judge_global_step_50_actor_huggingface_hotpotqa_output_20251121035628_1_in_1.jsonConfiguration427.6 KB —
constructed_kg/hotpotqa_output/kg_extraction/_data_httsangaj_autograph_checkpoints_ablation_judge_qwen7b-1.5b-judge_global_step_50_actor_huggingface_hotpotqa_output_20251121040302_1_in_1.jsonConfiguration628.2 KB —
constructed_kg/hotpotqa_output/kg_extraction/_data_httsangaj_autograph_checkpoints_ablation_judge_qwen7b-1.5b-judge_global_step_50_actor_huggingface_hotpotqa_output_20251121041017_1_in_1.jsonConfiguration29.6 MB f939819d0a2d
constructed_kg/musique_output/kg_extraction/_root_data_checkpoints_20251120_123621_qwen2.5-7b-judge-1.5B-autograph-distract_easy-docsize15-textlinkingFalse-f1False_global_step_50_actor_huggingface_musique_output_20251120171554_1_in_1.jsonConfiguration3.9 MB —
constructed_kg/musique_output/kg_extraction/_root_data_checkpoints_20251120_123621_qwen2.5-7b-judge-1.5B-autograph-distract_easy-docsize15-textlinkingFalse-f1False_global_step_50_actor_huggingface_musique_output_20251120172524_1_in_1.jsonConfiguration393.4 KB —
constructed_kg/musique_output/kg_extraction/_root_data_checkpoints_20251120_123621_qwen2.5-7b-judge-1.5B-autograph-distract_easy-docsize15-textlinkingFalse-f1False_global_step_50_actor_huggingface_musique_output_20251120172702_1_in_1.jsonConfiguration215.3 KB —
constructed_kg/musique_output/kg_extraction/_root_data_checkpoints_20251120_123621_qwen2.5-7b-judge-1.5B-autograph-distract_easy-docsize15-textlinkingFalse-f1False_global_step_50_actor_huggingface_musique_output_20251120172803_1_in_1.jsonConfiguration34.2 MB 17a0e687c5d1
generation_config.jsonConfiguration243 B —
model.safetensors.index.jsonConfiguration27.8 KB —
special_tokens_map.jsonConfiguration613 B —
benchmark/graph_retrieval/2wikimultihopqa_1763701513.7159789_benchmark.logOther843 B —
benchmark/graph_retrieval/2wikimultihopqa_1763701998.6973448_benchmark.logOther843 B —
benchmark/graph_retrieval/2wikimultihopqa_1763702076.2732604_benchmark.logOther15.5 KB —
benchmark/graph_retrieval/2wikimultihopqa_1763702198.7103505_benchmark.logOther888 B —
benchmark/graph_retrieval/2wikimultihopqa_1763702234.1501956_benchmark.logOther888 B —
benchmark/graph_retrieval/2wikimultihopqa_1763702428.2084653_benchmark.logOther19.5 MB 230b71ae4475
benchmark/graph_retrieval/hotpotqa_1763730249.4176042_benchmark.logOther10.6 MB 38d3c6b73001
benchmark/graph_retrieval/musique_1763715976.1551065_benchmark.logOther15.6 MB a099300647f4
benchmark/graph_retrieval/nq_1763742971.3977606_benchmark.logOther6.5 MB —
benchmark/graph_retrieval/popqa_1763754015.1567483_benchmark.logOther3.1 MB —
chat_template.jinjaOther2.5 KB —
constructed_kg/2021wiki_output/kg_graphml/2021wiki_graph.graphmlOther68.4 MB d096c869268c
constructed_kg/2021wiki_output/kg_graphml/2021wiki_without_concept.pklOther25.1 MB eb2b735dc91e
constructed_kg/2021wiki_output/precompute/2021wiki_Qwen3-Embedding-0.6B_text_embeddings.pklOther28.9 MB ab64618f1b6e
constructed_kg/2021wiki_output/precompute/2021wiki_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_embeddings.pklOther347.8 MB b8b0b466c509
constructed_kg/2021wiki_output/precompute/2021wiki_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_faiss.indexOther389.4 MB 8f8063cb336e
constructed_kg/2021wiki_output/precompute/2021wiki_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_embeddings.pklOther320.8 MB ccb8162db7de
constructed_kg/2021wiki_output/precompute/2021wiki_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_faiss.indexOther359.1 MB 82cd1aedf0ac
constructed_kg/2021wiki_output/precompute/2021wiki_eventFalse_conceptFalse_edge_list.pklOther7.4 MB 055586c2c00c
constructed_kg/2021wiki_output/precompute/2021wiki_eventFalse_conceptFalse_node_list.pklOther5.2 MB 2a37d1d72c88
constructed_kg/2021wiki_output/precompute/2021wiki_original_text_dict_with_node_id.pklOther5.4 MB f78dd2c360e2
constructed_kg/2021wiki_output/precompute/2021wiki_text_faiss.indexOther32.4 MB 8d5ae0c51193
constructed_kg/2021wiki_output/precompute/2021wiki_text_list.pklOther4.9 MB 2cf8ba65ebe8
constructed_kg/2021wiki_output/triples_csv/missing_concepts_2021wiki_from_json.csvOther4.0 MB —
constructed_kg/2021wiki_output/triples_csv/text_edges_2021wiki_from_json.csvOther8.0 MB —
constructed_kg/2021wiki_output/triples_csv/text_nodes_2021wiki_from_json.csvOther5.4 MB —
constructed_kg/2021wiki_output/triples_csv/triple_edges_2021wiki_from_json_without_emb.csvOther6.8 MB —
constructed_kg/2021wiki_output/triples_csv/triple_nodes_2021wiki_from_json_without_emb.csvOther3.8 MB —
constructed_kg/2wikimultihopqa_output/kg_graphml/2wikimultihopqa_graph.graphmlOther41.7 MB 9451e924ac73
constructed_kg/2wikimultihopqa_output/kg_graphml/2wikimultihopqa_without_concept.pklOther16.4 MB 803f77d9be43
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_Qwen3-Embedding-0.6B_text_embeddings.pklOther25.3 MB 91b3818d3342
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_embeddings.pklOther233.0 MB ad97d09a18a7
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_faiss.indexOther260.9 MB 4206a77c9927
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_embeddings.pklOther208.2 MB 617dddeadb0e
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_faiss.indexOther233.1 MB 3900c56afa0a
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_eventFalse_conceptFalse_edge_list.pklOther5.0 MB 728d10e42a38
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_eventFalse_conceptFalse_node_list.pklOther3.4 MB 1f5c0dbd78ef
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_original_text_dict_with_node_id.pklOther3.2 MB 63892500be9b
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_text_faiss.indexOther28.3 MB 500c0b4b0962
constructed_kg/2wikimultihopqa_output/precompute/2wikimultihopqa_text_list.pklOther2.8 MB 1b89607b6103
constructed_kg/2wikimultihopqa_output/triples_csv/missing_concepts_2wikimultihopqa_from_json.csvOther2.2 MB —
constructed_kg/2wikimultihopqa_output/triples_csv/text_edges_2wikimultihopqa_from_json.csvOther5.0 MB —
constructed_kg/2wikimultihopqa_output/triples_csv/text_nodes_2wikimultihopqa_from_json.csvOther3.2 MB —
constructed_kg/2wikimultihopqa_output/triples_csv/triple_edges_2wikimultihopqa_from_json_without_emb.csvOther4.3 MB —
constructed_kg/2wikimultihopqa_output/triples_csv/triple_nodes_2wikimultihopqa_from_json_without_emb.csvOther2.3 MB —
constructed_kg/hotpotqa_output/kg_graphml/hotpotqa_graph.graphmlOther81.0 MB 8766c5108018
constructed_kg/hotpotqa_output/kg_graphml/hotpotqa_without_concept.pklOther30.1 MB 7d198482cd10
constructed_kg/hotpotqa_output/precompute/hotpotqa_Qwen3-Embedding-0.6B_text_embeddings.pklOther38.1 MB c6f7e6db3ec9
constructed_kg/hotpotqa_output/precompute/hotpotqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_embeddings.pklOther429.1 MB d769a6ebd2bd
constructed_kg/hotpotqa_output/precompute/hotpotqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_faiss.indexOther480.3 MB d3c6f043b4b8
constructed_kg/hotpotqa_output/precompute/hotpotqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_embeddings.pklOther375.9 MB 20ed1a83479c
constructed_kg/hotpotqa_output/precompute/hotpotqa_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_faiss.indexOther420.8 MB 7866dd1ce541
constructed_kg/hotpotqa_output/precompute/hotpotqa_eventFalse_conceptFalse_edge_list.pklOther8.9 MB f7853281b4ba
constructed_kg/hotpotqa_output/precompute/hotpotqa_eventFalse_conceptFalse_node_list.pklOther6.1 MB fd060e709d9a
constructed_kg/hotpotqa_output/precompute/hotpotqa_original_text_dict_with_node_id.pklOther5.9 MB 2106efc153e9
constructed_kg/hotpotqa_output/precompute/hotpotqa_text_faiss.indexOther42.6 MB b71197ffc238
constructed_kg/hotpotqa_output/precompute/hotpotqa_text_list.pklOther5.3 MB 7a569a928085
constructed_kg/hotpotqa_output/triples_csv/missing_concepts_hotpotqa_from_json.csvOther4.2 MB —
constructed_kg/hotpotqa_output/triples_csv/text_edges_hotpotqa_from_json.csvOther9.2 MB —
constructed_kg/hotpotqa_output/triples_csv/text_nodes_hotpotqa_from_json.csvOther6.0 MB —
constructed_kg/hotpotqa_output/triples_csv/triple_edges_hotpotqa_from_json_without_emb.csvOther8.1 MB —
constructed_kg/hotpotqa_output/triples_csv/triple_nodes_hotpotqa_from_json_without_emb.csvOther4.3 MB —
constructed_kg/musique_output/kg_graphml/musique_graph.graphmlOther91.2 MB ba72426f5082
constructed_kg/musique_output/kg_graphml/musique_without_concept.pklOther36.4 MB 24dd398a201f
constructed_kg/musique_output/precompute/musique_Qwen3-Embedding-0.6B_text_embeddings.pklOther48.1 MB 39ef5a8ea2af
constructed_kg/musique_output/precompute/musique_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_embeddings.pklOther522.9 MB 20becc994c58
constructed_kg/musique_output/precompute/musique_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_edge_faiss.indexOther585.4 MB d729d4570cb0
constructed_kg/musique_output/precompute/musique_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_embeddings.pklOther451.7 MB 47335982e657
constructed_kg/musique_output/precompute/musique_eventFalse_conceptFalse_Qwen3-Embedding-0.6B_node_faiss.indexOther505.7 MB f8aea7b6fe35
constructed_kg/musique_output/precompute/musique_eventFalse_conceptFalse_edge_list.pklOther11.1 MB 8cbf5e35783a
constructed_kg/musique_output/precompute/musique_eventFalse_conceptFalse_node_list.pklOther7.3 MB a37dc89eee4b
constructed_kg/musique_output/precompute/musique_original_text_dict_with_node_id.pklOther6.7 MB eaa5e6e4c83f
constructed_kg/musique_output/precompute/musique_text_faiss.indexOther53.9 MB 3b18aa1a8e14
constructed_kg/musique_output/precompute/musique_text_list.pklOther6.0 MB 0e9164405c42
constructed_kg/musique_output/triples_csv/missing_concepts_musique_from_json.csvOther5.0 MB —
constructed_kg/musique_output/triples_csv/text_edges_musique_from_json.csvOther11.0 MB c1f4f24cb876
constructed_kg/musique_output/triples_csv/text_nodes_musique_from_json.csvOther6.8 MB —
constructed_kg/musique_output/triples_csv/triple_edges_musique_from_json_without_emb.csvOther10.2 MB —
constructed_kg/musique_output/triples_csv/triple_nodes_musique_from_json_without_emb.csvOther5.0 MB —
.gitattributesRepository5.1 KB —
merges.txtTokenizer1.7 MB —
tokenizer.jsonTokenizer11.4 MB 9c5ae00e602b
tokenizer_config.jsonTokenizer4.7 KB —
vocab.jsonTokenizer2.8 MB —

License and Download

License
Not stated by the source
Access
Open weights, no gate
Download size
15.2 GB
Download from Hong Ting Tsang

Released by Hong Ting Tsang through its official repository on Hugging Face.

Memory Requirements

PrecisionWeights in memory
As published15.2 GB
16-bit15.2 GB
8-bit7.6 GB
4-bit3.8 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b

How much GPU memory does AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b need?

About 18.3 GB at 16-bit and 4.6 GB at 4-bit: the weights (7.6B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

What is AutoGraphR1-musique_hotpotqa_train-qwen2.5-7b-graph-retriever-grpo-judge-1.5b's context length?

32,768 tokens, from the maximum position embeddings in its published configuration.