Ricardo Malagon Jerez

rjmalagon

AI & ML interests

Local inferencing: low-cost and high-precision models inferencing in cheap common hardware. Resilience, data sovereignty, private on-premise inferencing and agentic infrastructure. Quality model testing for OSS and OpenWeights models. Implementing highly specialized small LLM models (medicine, education, industry engineering, etc.) in community-owned servers in small public universities, small and remote access towns.

Recent Activity

new activity 27 days ago

Zaynoid/JSL-VL-7B-V1.2:Asking permission to access the model, and a proper way to share observations with you.

liked a model 9 months ago

Danielbrdz/Barcenas-Qwen3-14B

liked a model 9 months ago

QuixiAI/Qwen3-72B-Synthesis

View all activity

Organizations

None yet

New activity in Zaynoid/JSL-VL-7B-V1.2 27 days ago

Asking permission to access the model, and a proper way to share observations with you.

#1 opened 27 days ago by

rjmalagon

liked 2 models 9 months ago

Danielbrdz/Barcenas-Qwen3-14B

Text Generation • 15B • Updated Jun 9, 2025 • 3 • 3

QuixiAI/Qwen3-72B-Synthesis

73B • Updated Jun 11, 2025 • 106 • 8

liked a Space 10 months ago

MEDIC Benchmark

📊

Explore LLM benchmark leaderboards and results

liked 2 models about 1 year ago

wanlige/li-14b-v0.4

Text Generation • 15B • Updated Apr 30, 2025 • 132 • • 21

Danielbrdz/Barcenas-10b

Text Generation • 10B • Updated Jan 4, 2025 • 5 • 1

reacted to mkurman's post with 🔥 over 1 year ago

Post

1190

We built a new small language model SmolLM2-MedIT-Upscale-2B, based on SmolLM2-1.7B-Instruct from Hugging Face. The premise was simple - increasing the vector in attention layers would positively impact the model's capabilities.

What did we prove?
In total, not much really, since we don't have the original trained under the same conditions as our upscale. However...

1. We scaled up the model without losing its quality
2. We confirmed that the method we devised works
3. After extremely short fine-tuning, the model achieved much better results in IFEval compared to the original (53.68 vs 64.29) and a higher overall average score in Open LLM Leaderboard (14.75 vs 15.17)

I consider this a big success 😇, since surpassing the original in metrics is often very time-consuming, generates high costs, and doesn't always work out.

Meanwhile, we're moving forward, training SmolLM2 400M Instruct as an upscale of 136M.

We're curious about how increasing the base and intermediate vectors will affect the model's quality. We'll compare it to the original and the 360M Instruct version released by Hugging Face.

License: Apache 2.0

meditsolutions/SmolLM2-MedIT-Upscale-2B

reacted to clem's post with 👍 over 1 year ago

Post

4606

Hugging Face is becoming the best place to share the most viral AI apps with spaces.

Kolors Virtual Try-on just crossed 6,000,000 unique visitors & is now the #5 most popular space. Congrats to the Kwai Kolors team!

Kwai-Kolors/Kolors-Virtual-Try-On