aethertp/PicoLM-V2-81M-Instruct
Text Generation • 96M • Updated • 509 • 2
Aether
aethertp
AI & ML interests
None yet
Recent Activity
updated a model 3 days ago
aethertp/PicoLM-V2-81M-InstructOrganizations
None yet
Drop redundant untied lm_head.weight (Option 1) The checkpoint stored both tok_embeddings.weight and a separate lm_head.weight, each [24576, 576] = 14,155,776 params, even though config.json sets tie_word_embeddings: true. At runtime the head is tied to the embedding, so lm_head.weight is never read. This removes that one tensor. Verified: - 201 -> 200 tensors; stored elements 96,017,472 -> 81,861,696 (exactly the card's count) - file 192,054,704 -> 163,743,061 bytes (-28.3 MB) - loaded both versions with the repo's own PicoLMV2ForCausalLM: identical logits on fixed inputs (max |diff| = 0.0), same 81,861,696 unique params, head tied to embedding in both. Requested by aethertp in the discussion (Option 1).
#2 opened 2 days ago
by 
Compactbot
Checkpoint stores an untied lm_head: 96.0M params on disk vs 81.86M on the card
4
#1 opened 3 days ago
by Compactbot
updated a
model 3 days ago
Overcoming the ARC-Easy floor and vocabulary budget on an 80M footprint
❤️ 1
8
#1 opened 7 days ago
by 
AndrewThompson1233
liked 2
models 4 days ago
updated a
model 4 days ago
published a
model 4 days ago
liked a
model 4 days ago
updated a
model 4 days ago
liked a
model 4 days ago
published a
model 4 days ago
published a
model 7 days ago
liked a
Space 3 months ago
liked a
model 7 months ago
published a
Space about 1 year ago