jxm/gpt-oss-20b-base
Text Generation
•
21B
•
Updated
•
1.33k
•
140
sometimesanotion
sometimesanotion
AI & ML interests
Agentic LLM services, model merging, finetunes, distillation
Recent Activity
liked
a model
about 8 hours ago
jxm/gpt-oss-20b-base
upvoted
a
collection
about 12 hours ago
GPT-OSS Pruned Experts (4.2B-20B) [IF, Science, Math, etc.]
liked
a dataset
about 13 hours ago
AmanPriyanshu/GPT-OSS-20B-MoE-expert-activations
Organizations
liked
a
model
about 8 hours ago
upvoted
a
collection
about 12 hours ago
liked
a
dataset
about 13 hours ago
upvoted
a
collection
about 13 hours ago
liked
2
models
about 13 hours ago
liked
a
dataset
9 days ago
I also have a hypothesis that this model can be efficiently downsized not by pruning experts, but by using merges and LoRAs to downsize their unique parameter count. The merge would be most of the shared parameters, and the routing table need not change.
I'm building up a new version of my pipeline to test this hypothesis. I suspect it'd let us get most of the performance in <12B parameters.
This is a very cool release! I really enjoy the ShiningValiant series!
Do you see potential to prune experts or layers from the gpt-oss-20b model to downsize it, and then finetune?
reacted to
sequelbox's
post with 🔥👍
10 days ago
Post
2161
NEW RELEASE: Shining Valiant 3 now available for openai/gpt-oss-20b!
- Cutting edge science-reasoning: sequelbox/Celestia3-DeepSeek-R1-0528 for physics, biology, chemistry, compsci, astronomy, Earth science, and information theory.
- AI to build AI: the all-new sequelbox/Mitakihara-DeepSeek-R1-0528 dataset for high-quality reasoning performance on AI, MLOps, math and CUDA, complex adaptive and agentic systems, cognition, logic, linguistics, simulation, knowledge management, and more!
- Creative reasoning and general chat performance supplemented with sequelbox/Raiden-DeepSeek-R1
Get the new SV3: ValiantLabs/gpt-oss-20b-ShiningValiant3
This is our first release for the new openai/gpt-oss-20b - we're hoping to support this model with more releases going forward.
We're also excited to bring our models to Qwen/Qwen3-4B-Thinking-2507 and the other 2507 Qwen 3 models - coming very soon!
We want to bring SV3, Esper 3, and our Experimental Reasoning finetunes to more models ASAP. Help us out: sequelbox/SupportOpenSource
Open source matters. Fight for it with us.
love,
allegra
- Cutting edge science-reasoning: sequelbox/Celestia3-DeepSeek-R1-0528 for physics, biology, chemistry, compsci, astronomy, Earth science, and information theory.
- AI to build AI: the all-new sequelbox/Mitakihara-DeepSeek-R1-0528 dataset for high-quality reasoning performance on AI, MLOps, math and CUDA, complex adaptive and agentic systems, cognition, logic, linguistics, simulation, knowledge management, and more!
- Creative reasoning and general chat performance supplemented with sequelbox/Raiden-DeepSeek-R1
Get the new SV3: ValiantLabs/gpt-oss-20b-ShiningValiant3
This is our first release for the new openai/gpt-oss-20b - we're hoping to support this model with more releases going forward.
We're also excited to bring our models to Qwen/Qwen3-4B-Thinking-2507 and the other 2507 Qwen 3 models - coming very soon!
We want to bring SV3, Esper 3, and our Experimental Reasoning finetunes to more models ASAP. Help us out: sequelbox/SupportOpenSource
Open source matters. Fight for it with us.
love,
allegra
-
- 3 replies
liked
a
model
10 days ago
reacted to
sequelbox's
post with 🚀
19 days ago
Post
2289
NEW EXPERIMENTAL RELEASE: DAG Reasoning is here!
- Our first Experimental Reasoning Modality release: create structured, analytical Directed Acyclic Graphs to provide insight into your queries and situations!
- Multi-step analysis identifies causal relationships, produces confidence measurements, and forms a single structured graph object.
- DAG Reasoning Format provides clear, readable JSON containing structured, useful information; easy to use for creating visualizations, doing analysis, or further conversation with your assistant.
- Trained in a variety of subjects for flexible analysis: programming, science, business, economics, finance, law, logistics, management, and more!
Our first DAG Reasoning release is Qwen 3, starting off with 8B and 14B!
Get 8B: sequelbox/Qwen3-8B-DAG-Reasoning
Get 14B: sequelbox/Qwen3-14B-DAG-Reasoning
You can also get the DAG Reasoning dataset, to train your own models to use DAG Reasoning Format: sequelbox/DAG-Reasoning-DeepSeek-R1-0528
Support our experimental open-source research efforts, models and datasets: sequelbox/SupportOpenSource
with love,
allegra
- Our first Experimental Reasoning Modality release: create structured, analytical Directed Acyclic Graphs to provide insight into your queries and situations!
- Multi-step analysis identifies causal relationships, produces confidence measurements, and forms a single structured graph object.
- DAG Reasoning Format provides clear, readable JSON containing structured, useful information; easy to use for creating visualizations, doing analysis, or further conversation with your assistant.
- Trained in a variety of subjects for flexible analysis: programming, science, business, economics, finance, law, logistics, management, and more!
Our first DAG Reasoning release is Qwen 3, starting off with 8B and 14B!
Get 8B: sequelbox/Qwen3-8B-DAG-Reasoning
Get 14B: sequelbox/Qwen3-14B-DAG-Reasoning
You can also get the DAG Reasoning dataset, to train your own models to use DAG Reasoning Format: sequelbox/DAG-Reasoning-DeepSeek-R1-0528
Support our experimental open-source research efforts, models and datasets: sequelbox/SupportOpenSource
with love,
allegra
-
- 2 replies
liked
2
models
23 days ago
liked
2
models
24 days ago
liked
a
model
25 days ago
reacted to
sequelbox's
post with 🔥
about 1 month ago
Post
2605
Some new releases:
- brought the new Shining Valiant 3 series (science-reasoning, AI-reasoning, general chat) to Qwen 3 4B: ValiantLabs/Qwen3-4B-ShiningValiant3
- merged models for Shining Valiant 3 and Esper 3, combining their technical expertise and reasoning skills:
4b: sequelbox/Qwen3-4B-PlumEsper
8b: sequelbox/Qwen3-8B-PlumEsper
coming up we'll have some experimental reasoning releases - datasets and models will be out soon!
also will be bringing SV3 and Esper 3 to more models.
lets keep working for open source :)
love,
allegra
- brought the new Shining Valiant 3 series (science-reasoning, AI-reasoning, general chat) to Qwen 3 4B: ValiantLabs/Qwen3-4B-ShiningValiant3
- merged models for Shining Valiant 3 and Esper 3, combining their technical expertise and reasoning skills:
4b: sequelbox/Qwen3-4B-PlumEsper
8b: sequelbox/Qwen3-8B-PlumEsper
coming up we'll have some experimental reasoning releases - datasets and models will be out soon!
also will be bringing SV3 and Esper 3 to more models.
lets keep working for open source :)
love,
allegra
liked
a
model
about 1 month ago
