Epoch 2. Batch Size 0. Peers 10.
Browse files- config.json +11 -5
- inner_optimizer.pt +1 -1
- model.safetensors +1 -1
- outer_optimizer.pt +1 -1
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
@@ -160,7 +160,7 @@
|
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
"24": "NON_PARTICIPATING",
|
163 |
-
"240": "
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
"243": "NON_PARTICIPATING",
|
@@ -169,7 +169,7 @@
|
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
-
"249": "
|
173 |
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
175 |
"251": "NON_PARTICIPATING",
|
@@ -267,7 +267,13 @@
|
|
267 |
"AutoConfig": "distributed/optimized-gpt2-500m--configuration_gpt_optimized.GPTOptimConfig",
|
268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
-
"block_list": [
|
|
|
|
|
|
|
|
|
|
|
|
|
271 |
"block_size": 1024,
|
272 |
"bos_token_id": 50256,
|
273 |
"embd_pdrop": 0.1,
|
@@ -275,7 +281,7 @@
|
|
275 |
"initializer_range": 0.02,
|
276 |
"inner_step": 0,
|
277 |
"inner_steps": 0,
|
278 |
-
"last_allreduce_block":
|
279 |
"layer_norm_epsilon": 1e-05,
|
280 |
"model_type": "gpt_optimized",
|
281 |
"n_embd": 1280,
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "Mhair/onetwothree",
|
3 |
"activation_function": "gelu_new",
|
4 |
"all_reduce_scores": {
|
5 |
"0": "NON_PARTICIPATING",
|
|
|
160 |
"238": "NON_PARTICIPATING",
|
161 |
"239": "NON_PARTICIPATING",
|
162 |
"24": "NON_PARTICIPATING",
|
163 |
+
"240": "SUCCESS",
|
164 |
"241": "NON_PARTICIPATING",
|
165 |
"242": "NON_PARTICIPATING",
|
166 |
"243": "NON_PARTICIPATING",
|
|
|
169 |
"246": "NON_PARTICIPATING",
|
170 |
"247": "NON_PARTICIPATING",
|
171 |
"248": "NON_PARTICIPATING",
|
172 |
+
"249": "NON_PARTICIPATING",
|
173 |
"25": "SUCCESS",
|
174 |
"250": "NON_PARTICIPATING",
|
175 |
"251": "NON_PARTICIPATING",
|
|
|
267 |
"AutoConfig": "distributed/optimized-gpt2-500m--configuration_gpt_optimized.GPTOptimConfig",
|
268 |
"AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
|
269 |
},
|
270 |
+
"block_list": [
|
271 |
+
5064009,
|
272 |
+
5064013,
|
273 |
+
5064017,
|
274 |
+
5064021,
|
275 |
+
5064025
|
276 |
+
],
|
277 |
"block_size": 1024,
|
278 |
"bos_token_id": 50256,
|
279 |
"embd_pdrop": 0.1,
|
|
|
281 |
"initializer_range": 0.02,
|
282 |
"inner_step": 0,
|
283 |
"inner_steps": 0,
|
284 |
+
"last_allreduce_block": 5265017,
|
285 |
"layer_norm_epsilon": 1e-05,
|
286 |
"model_type": "gpt_optimized",
|
287 |
"n_embd": 1280,
|
inner_optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 2752
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b5f7b91e591ec147fc34a4b0d645772daee4b68d91c1416dff8233984a37df72
|
3 |
size 2752
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4040701744
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8be83e1ea9c0eaeab2cecfdeae6e9617dae603fdfbeda43c55bc385455934a1f
|
3 |
size 4040701744
|
outer_optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4040805354
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1dd3daddc2a7543570712cda9cc14833d6f09a833e09a76eb9f5a1ac0ba93eb2
|
3 |
size 4040805354
|