suitch commited on
Commit
54f9ce8
·
verified ·
1 Parent(s): 032c3ef

Point to thesis repository

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -14,7 +14,7 @@ tags:
14
 
15
  # DPO-tuned report decoder
16
 
17
- Part of the master's thesis *From Scan to Report: Detecting, Segmenting and Reporting Findings in Chest CT* (code: https://github.com/Switchharsh/From_Scan_to_Report_Thesis).
18
 
19
  The baseline decoder (`scan2report-decoder-baseline`) further tuned with DPO on 1,899 preference pairs (beta 0.1, lr 2e-5, 2 epochs).
20
 
@@ -29,7 +29,7 @@ Three-seed mean (each DPO trained from its own baseline): CRG 0.3965 +- 0.0139,
29
  - Input: the 8x8x8 FlexiCT token grid of a chest CT (`flexict/tokens8`, 864-dim per token) from CT-RATE.
30
  - `projector.pt`: MLP (864 -> 2560 -> 2560, LayerNorm) that maps the visual tokens into the LLM embedding space.
31
  - `adapter_model.safetensors` + `adapter_config.json`: LoRA (r=16, alpha=32) on [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B).
32
- - Loading and generation code (model wrapper, data, evaluation): https://github.com/Switchharsh/From_Scan_to_Report_Thesis (`ctbench/rgen/model.py`, `evaluate.py`). The adapter is not a stand-alone chat model: the projector and the FlexiCT tokens are required.
33
  - Task instruction used in every run: `Write the findings and impression for this chest CT.` Greedy decoding, at most 512 new tokens.
34
 
35
  ## Evaluation protocol
 
14
 
15
  # DPO-tuned report decoder
16
 
17
+ Part of the master's thesis *From Scan to Report: Detecting, Segmenting and Reporting Findings in Chest CT* (code: https://github.com/Switchharsh/Switchharsh-From_Scan_to_Report_Master_Thesis; thesis: https://github.com/Switchharsh/Switchharsh-From_Scan_to_Report_Master_Thesis/tree/main/thesis).
18
 
19
  The baseline decoder (`scan2report-decoder-baseline`) further tuned with DPO on 1,899 preference pairs (beta 0.1, lr 2e-5, 2 epochs).
20
 
 
29
  - Input: the 8x8x8 FlexiCT token grid of a chest CT (`flexict/tokens8`, 864-dim per token) from CT-RATE.
30
  - `projector.pt`: MLP (864 -> 2560 -> 2560, LayerNorm) that maps the visual tokens into the LLM embedding space.
31
  - `adapter_model.safetensors` + `adapter_config.json`: LoRA (r=16, alpha=32) on [Qwen/Qwen3.5-4B](https://huggingface.co/Qwen/Qwen3.5-4B).
32
+ - Loading and generation code (model wrapper, data, evaluation): https://github.com/Switchharsh/Switchharsh-From_Scan_to_Report_Master_Thesis (`ctbench/rgen/model.py`, `evaluate.py`). The adapter is not a stand-alone chat model: the projector and the FlexiCT tokens are required.
33
  - Task instruction used in every run: `Write the findings and impression for this chest CT.` Greedy decoding, at most 512 new tokens.
34
 
35
  ## Evaluation protocol