Training in progress, step 650
Browse files
adapter_config.json
CHANGED
|
@@ -16,7 +16,7 @@
|
|
| 16 |
"layers_pattern": null,
|
| 17 |
"layers_to_transform": null,
|
| 18 |
"loftq_config": {},
|
| 19 |
-
"lora_alpha":
|
| 20 |
"lora_bias": false,
|
| 21 |
"lora_dropout": 0.05,
|
| 22 |
"megatron_config": null,
|
|
@@ -24,39 +24,39 @@
|
|
| 24 |
"modules_to_save": null,
|
| 25 |
"peft_type": "LORA",
|
| 26 |
"qalora_group_size": 16,
|
| 27 |
-
"r":
|
| 28 |
"rank_pattern": {},
|
| 29 |
"revision": null,
|
| 30 |
"target_modules": [
|
| 31 |
-
"projections.16",
|
| 32 |
-
"key",
|
| 33 |
-
"text_projection",
|
| 34 |
-
"projections.12",
|
| 35 |
"qkv",
|
| 36 |
-
"
|
|
|
|
|
|
|
| 37 |
"projections.11",
|
| 38 |
-
"projections.
|
| 39 |
-
"projections.
|
| 40 |
-
"projections.
|
|
|
|
|
|
|
|
|
|
| 41 |
"value",
|
| 42 |
-
"projections.3",
|
| 43 |
-
"vision_projection",
|
| 44 |
-
"projections.0",
|
| 45 |
-
"itm_head",
|
| 46 |
-
"projections.4",
|
| 47 |
-
"projection",
|
| 48 |
"projections.10",
|
| 49 |
"projections.6",
|
| 50 |
"projections.5",
|
|
|
|
|
|
|
|
|
|
| 51 |
"dense",
|
|
|
|
| 52 |
"projections.9",
|
| 53 |
-
"projections.
|
| 54 |
-
"projections.
|
| 55 |
"fc1",
|
| 56 |
-
"
|
| 57 |
-
"projections.
|
| 58 |
-
"projections.
|
| 59 |
-
"projections.
|
|
|
|
| 60 |
],
|
| 61 |
"target_parameters": null,
|
| 62 |
"task_type": null,
|
|
|
|
| 16 |
"layers_pattern": null,
|
| 17 |
"layers_to_transform": null,
|
| 18 |
"loftq_config": {},
|
| 19 |
+
"lora_alpha": 32,
|
| 20 |
"lora_bias": false,
|
| 21 |
"lora_dropout": 0.05,
|
| 22 |
"megatron_config": null,
|
|
|
|
| 24 |
"modules_to_save": null,
|
| 25 |
"peft_type": "LORA",
|
| 26 |
"qalora_group_size": 16,
|
| 27 |
+
"r": 32,
|
| 28 |
"rank_pattern": {},
|
| 29 |
"revision": null,
|
| 30 |
"target_modules": [
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
"qkv",
|
| 32 |
+
"key",
|
| 33 |
+
"projections.4",
|
| 34 |
+
"projections.8",
|
| 35 |
"projections.11",
|
| 36 |
+
"projections.1",
|
| 37 |
+
"projections.2",
|
| 38 |
+
"projections.16",
|
| 39 |
+
"fc2",
|
| 40 |
+
"text_projection",
|
| 41 |
+
"projections.15",
|
| 42 |
"value",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
"projections.10",
|
| 44 |
"projections.6",
|
| 45 |
"projections.5",
|
| 46 |
+
"projections.17",
|
| 47 |
+
"projection",
|
| 48 |
+
"vision_projection",
|
| 49 |
"dense",
|
| 50 |
+
"itm_head",
|
| 51 |
"projections.9",
|
| 52 |
+
"projections.7",
|
| 53 |
+
"projections.12",
|
| 54 |
"fc1",
|
| 55 |
+
"projections.13",
|
| 56 |
+
"projections.3",
|
| 57 |
+
"projections.0",
|
| 58 |
+
"projections.14",
|
| 59 |
+
"query"
|
| 60 |
],
|
| 61 |
"target_parameters": null,
|
| 62 |
"task_type": null,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3290475cbd1c0116ceab9196965acc1bcb5f5d6624bb9582cf8459626d021ecc
|
| 3 |
+
size 157934960
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4773786592
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10fb7c973c4a255c34f3e4a5f0541867af088563429f4a43652e84d8262f864f
|
| 3 |
size 4773786592
|
runs/Oct08_08-21-55_e41bbb0dd69c/events.out.tfevents.1759911716.e41bbb0dd69c.134725.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99b66cb0745257ee7bd92b15633409a1ce03589d9835efc07657dd261ee70e94
|
| 3 |
+
size 145436
|