sesame/csm-1b
Text-to-Speech
•
Updated
•
61.9k
•
•
1.73k
Siddish Iragamreddy
Siddish
AI & ML interests
None yet
Recent Activity
liked
a model
17 days ago
sesame/csm-1b
upvoted
a
collection
about 1 month ago
Foundation Text-Generation Models Below 360M Parameters
liked
a model
about 1 month ago
HuggingFaceTB/SmolLM2-360M
Organizations
Siddish's activity
liked
a
model
17 days ago
upvoted
a
collection
about 1 month ago
liked
a
model
about 1 month ago
liked
3
datasets
about 2 months ago
upvoted
an
article
about 2 months ago
Article
DABStep: Data Agent Benchmark for Multi-step Reasoning
•
67
liked
7
datasets
about 2 months ago
proj-persona/PersonaHub
Viewer
•
Updated
•
375k
•
11.1k
•
549
xywang1/OpenCharacter
Viewer
•
Updated
•
326k
•
203
•
15
gretelai/synthetic_text_to_sql
Viewer
•
Updated
•
106k
•
6.89k
•
512
TIGER-Lab/AceCode-87K
Viewer
•
Updated
•
87.1k
•
789
•
37
open-thoughts/OpenThoughts-114k
Viewer
•
Updated
•
228k
•
29.7k
•
673
ServiceNow-AI/R1-Distill-SFT
Viewer
•
Updated
•
1.85M
•
3.2k
•
290
lmms-lab/multimodal-open-r1-8k-verified
Viewer
•
Updated
•
7.69k
•
2.54k
•
49
reacted to
Ihor's
post with 🔥
about 2 months ago
Post
1538
🚀 Reproducing DeepSeek R1 for Text-to-Graph Extraction
I’ve been working on replicating DeepSeek R1, focusing on zero-shot text-to-graph extraction—a challenging task where LMs extract entities and relations from text based on predefined types.
🧠Key Insight:
Language models struggle when constrained by entity/relation types. Supervised training alone isn’t enough, but reinforcement learning (RL), specifically Guided Reward Policy Optimization (GRPO), shows promise.
💡 Why GRPO?
It trains the model to generate structured graphs, optimizing multiple reward functions (format, JSON validity, and extraction accuracy).
It allows the model to learn from both positive and hard negative examples dynamically.
RL can be fine-tuned to emphasize relation extraction improvements.
📊 Early Results:
Even with limited training, F1 scores consistently improved, and we saw clear benefits from RL-based optimization. More training = better performance!
🔬 Next Steps:
We’re scaling up experiments with larger models and high-quality data. Stay tuned for updates! Meanwhile, check out one of our experimental models here:
Ihor/Text2Graph-R1-Qwen2.5-0.5b
📔 Learn more details from the blog post: https://medium.com/p/d8b648d9f419
Feel free to share your thoughts and ask questions!
I’ve been working on replicating DeepSeek R1, focusing on zero-shot text-to-graph extraction—a challenging task where LMs extract entities and relations from text based on predefined types.
🧠Key Insight:
Language models struggle when constrained by entity/relation types. Supervised training alone isn’t enough, but reinforcement learning (RL), specifically Guided Reward Policy Optimization (GRPO), shows promise.
💡 Why GRPO?
It trains the model to generate structured graphs, optimizing multiple reward functions (format, JSON validity, and extraction accuracy).
It allows the model to learn from both positive and hard negative examples dynamically.
RL can be fine-tuned to emphasize relation extraction improvements.
📊 Early Results:
Even with limited training, F1 scores consistently improved, and we saw clear benefits from RL-based optimization. More training = better performance!
🔬 Next Steps:
We’re scaling up experiments with larger models and high-quality data. Stay tuned for updates! Meanwhile, check out one of our experimental models here:
Ihor/Text2Graph-R1-Qwen2.5-0.5b
📔 Learn more details from the blog post: https://medium.com/p/d8b648d9f419
Feel free to share your thoughts and ask questions!
- 2 replies
liked
2
models
about 2 months ago
liked
2
models
2 months ago
liked
a
model
4 months ago
