klein-ml/Klein-TTS-X1.5-Flash-Lite
Text-to-Speech • 401k • Updated • 36 • 5
🔄 In a Training Loop
Inflammable1230
Inflammable1230
AI & ML interests
Edge AI enthusiast, 16GB CPU, Falcon H1 90M = favorite SLM.
Recent Activity
liked a model 1 day ago
klein-ml/Klein-TTS-X1.5-Flash-Lite liked a model 3 days ago
ML-Intern-lab/Qwen3.5-0.8B-weight-space-programming liked a model 6 days ago
microsoft/UserLM-2.0Organizations
None yet
liked a
model 1 day ago
liked a
model 3 days ago
liked 2
models 6 days ago
reacted to crp4222's post with 🔥 6 days ago
Post
2,735
I trained a small PII detector for PrivAiTe, my local proxy that masks personal data before a prompt reaches an LLM: crp4222/privaite-pii-minilm.
It is a multilingual MiniLM (117M parameters, 15 types from names and contacts to IDs, bank details and secrets), trained only on datasets whose licenses allow commercial use, released under BSD 3-Clause, with PyTorch, ONNX fp32 and int8, and Core ML builds.
On 9 public test sets it never saw, it masks as many person names as the best of GLiNER PII, NVIDIA GLiNER-PII, GLiNER2, Piiranha and OpenAI privacy-filter, or more, on 6 of them, usually with higher precision. It trails GLiNER PII on short lowercase voice requests and NVIDIA on court rulings.
It is also 2.5 to 7 times smaller and much faster: 9 ms per 1,000 characters on an M1 Pro Neural Engine and 23 ms on CPU in int8, against 186 to 621 ms for the others on the same machine and documents. Numbers, test sets and limits are on the card: crp4222/privaite-pii-minilm
It is a multilingual MiniLM (117M parameters, 15 types from names and contacts to IDs, bank details and secrets), trained only on datasets whose licenses allow commercial use, released under BSD 3-Clause, with PyTorch, ONNX fp32 and int8, and Core ML builds.
On 9 public test sets it never saw, it masks as many person names as the best of GLiNER PII, NVIDIA GLiNER-PII, GLiNER2, Piiranha and OpenAI privacy-filter, or more, on 6 of them, usually with higher precision. It trails GLiNER PII on short lowercase voice requests and NVIDIA on court rulings.
It is also 2.5 to 7 times smaller and much faster: 9 ms per 1,000 characters on an M1 Pro Neural Engine and 23 ms on CPU in int8, against 186 to 621 ms for the others on the same machine and documents. Numbers, test sets and limits are on the card: crp4222/privaite-pii-minilm
- 5 replies
upvoted a paper 6 days ago
liked 2
models 7 days ago
liked 2
models 8 days ago
reacted to kaupane's post with 🚀 8 days ago
Post
2,054
I built Inko end to end: a 0.5B bilingual text-to-image model with a focus on fine art, from data curation and pretraining to RL post-training and evaluation.
It competes with DreamLite, SANA-0.6B and PixArt-Σ, and approaches SD3.5 Medium on some benchmarks.
The journey involves four rounds of cleaning the Chinese-painting dataset, a cross-hardware loss-spike investigation, and a reward redesign after early RL experiments collapased diversity.
Behind the release, a four-domain bilingual corpus, 600k pretraining updates on 16 Ascend NPUs, and DiffusionART with a custom 24.6k-prompt pool and five pairwise rubric channels.
Dense prompt alignment is Inko's strongest general result. Anatomy, counting, reasoning, text rendering, and exact spatial relations remain limitations.
kaupane/Inko
kaupane/Inko
It competes with DreamLite, SANA-0.6B and PixArt-Σ, and approaches SD3.5 Medium on some benchmarks.
The journey involves four rounds of cleaning the Chinese-painting dataset, a cross-hardware loss-spike investigation, and a reward redesign after early RL experiments collapased diversity.
Behind the release, a four-domain bilingual corpus, 600k pretraining updates on 16 Ascend NPUs, and DiffusionART with a custom 24.6k-prompt pool and five pairwise rubric channels.
Dense prompt alignment is Inko's strongest general result. Anatomy, counting, reasoning, text rendering, and exact spatial relations remain limitations.
kaupane/Inko
kaupane/Inko
- 1 reply
liked a
model 9 days ago
liked a
model 12 days ago
liked a
model 13 days ago
liked a
model 14 days ago
upvoted an article 14 days ago
Article
Laghu: a 152M chat model trained from scratch for $146. Everything is open.
udgamlabs
• • 3 liked a
Space 14 days ago
reacted to jiteshkprasad's post with 🚀 14 days ago
Post
1,740
Laghu: a 152M chat model trained from scratch for $146 of pretraining compute, released fully open. It thinks before it answers (on/off), holds multi-turn conversations and chats, in 163 MB that runs in your browser.
- Base model, new text (Sep-Oct 2026): 0.919 bits per byte vs SmolLM2-135M 0.941 and Falcon-H1-Tiny-90M-Base 0.979
- Chat: level with SmolLM2-135M-Instruct on chat quality and multi-turn, ahead on maths with thinking; behind Falcon-H1-Tiny-90M-Instruct on all three
- Thinking on/off, 16K context; runs in the browser, Ollama (
- Every surviving checkpoint with optimiser states, code, recipe, per-item evals, failures and cost
How we made it: https://huggingface.co/blog/udgamlabs/laghu
Collection: udgamlabs/laghu-6ad91b153ba8cdc4e40900bd
Space: udgamlabs/Laghu
- Base model, new text (Sep-Oct 2026): 0.919 bits per byte vs SmolLM2-135M 0.941 and Falcon-H1-Tiny-90M-Base 0.979
- Chat: level with SmolLM2-135M-Instruct on chat quality and multi-turn, ahead on maths with thinking; behind Falcon-H1-Tiny-90M-Instruct on all three
- Thinking on/off, 16K context; runs in the browser, Ollama (
ollama run hf.co/udgamlabs/Laghu-GGUF:Q8_0), LM Studio, llama.cpp, MLX and transformers- Every surviving checkpoint with optimiser states, code, recipe, per-item evals, failures and cost
How we made it: https://huggingface.co/blog/udgamlabs/laghu
Collection: udgamlabs/laghu-6ad91b153ba8cdc4e40900bd
Space: udgamlabs/Laghu
liked a
Space 14 days ago
liked a
model 14 days ago
