Upload folder using huggingface_hub
Browse files- v2/README.md +24 -0
- v2/adapter_config.json +46 -0
- v2/adapter_model.safetensors +3 -0
v2/README.md
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc0-1.0
|
| 3 |
+
base_model: Qwen/Qwen2.5-Coder-3B-Instruct
|
| 4 |
+
library_name: peft
|
| 5 |
+
tags: [lora, emergent-language, agent-communication, code, multi-task]
|
| 6 |
+
---
|
| 7 |
+
# Glyph v2 — multi-task LoRA (speaker · builder · translator in one adapter)
|
| 8 |
+
|
| 9 |
+
Finetuned on 40,592 pairs over a 68-primitive typed glyph language
|
| 10 |
+
(dataset: robertkeus/glyph-tasks, sft2.jsonl). One adapter, three roles by
|
| 11 |
+
instruction prompt:
|
| 12 |
+
|
| 13 |
+
- `Encode this task as glyph symbols.\nTask: {english}\nSymbols:`
|
| 14 |
+
- `Write Python for this glyph message.\nSymbols: {glyphs}\nCode:`
|
| 15 |
+
- `Translate this glyph message into English.\nSymbols: {glyphs}\nEnglish:`
|
| 16 |
+
|
| 17 |
+
## Results (seed 0, 3 epochs, Kaggle T4, LoRA r=32 all-linear)
|
| 18 |
+
| metric | score |
|
| 19 |
+
|---|---|
|
| 20 |
+
| builder: glyphs → correct code, HELD-OUT compositions (tests executed) | **99.5%** |
|
| 21 |
+
| speaker: UNSEEN phrasings → exact glyph message | **86.5%** |
|
| 22 |
+
| zero-shot symbols (never trained) | 2% |
|
| 23 |
+
|
| 24 |
+
Code + demo: https://github.com/robertkeus/glyph-ai
|
v2/adapter_config.json
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alora_invocation_tokens": null,
|
| 3 |
+
"alpha_pattern": {},
|
| 4 |
+
"arrow_config": null,
|
| 5 |
+
"auto_mapping": null,
|
| 6 |
+
"base_model_name_or_path": "Qwen/Qwen2.5-Coder-3B-Instruct",
|
| 7 |
+
"bias": "none",
|
| 8 |
+
"corda_config": null,
|
| 9 |
+
"ensure_weight_tying": false,
|
| 10 |
+
"eva_config": null,
|
| 11 |
+
"exclude_modules": null,
|
| 12 |
+
"fan_in_fan_out": false,
|
| 13 |
+
"inference_mode": true,
|
| 14 |
+
"init_lora_weights": true,
|
| 15 |
+
"layer_replication": null,
|
| 16 |
+
"layers_pattern": null,
|
| 17 |
+
"layers_to_transform": null,
|
| 18 |
+
"loftq_config": {},
|
| 19 |
+
"lora_alpha": 64,
|
| 20 |
+
"lora_bias": false,
|
| 21 |
+
"lora_dropout": 0.05,
|
| 22 |
+
"megatron_config": null,
|
| 23 |
+
"megatron_core": "megatron.core",
|
| 24 |
+
"modules_to_save": null,
|
| 25 |
+
"peft_type": "LORA",
|
| 26 |
+
"peft_version": "0.18.1",
|
| 27 |
+
"qalora_group_size": 16,
|
| 28 |
+
"r": 32,
|
| 29 |
+
"rank_pattern": {},
|
| 30 |
+
"revision": null,
|
| 31 |
+
"target_modules": [
|
| 32 |
+
"gate_proj",
|
| 33 |
+
"q_proj",
|
| 34 |
+
"v_proj",
|
| 35 |
+
"k_proj",
|
| 36 |
+
"down_proj",
|
| 37 |
+
"o_proj",
|
| 38 |
+
"up_proj"
|
| 39 |
+
],
|
| 40 |
+
"target_parameters": null,
|
| 41 |
+
"task_type": "CAUSAL_LM",
|
| 42 |
+
"trainable_token_indices": null,
|
| 43 |
+
"use_dora": false,
|
| 44 |
+
"use_qalora": false,
|
| 45 |
+
"use_rslora": false
|
| 46 |
+
}
|
v2/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bde2236d10d41836901625b5096db79797e3ace572bccad628693ff1cf1f331a
|
| 3 |
+
size 239536272
|