Merge pull request #105 from viktoriussuwandi/viktoriussuwandi-patch
Browse files- configs/cerebras_1_3B_alpaca.yml +1 -1
- configs/galactica_1_3B.yml +1 -1
- configs/gpt_neox_20b.yml +1 -1
- configs/llama_13B_alpaca.yml +1 -1
- configs/llama_65B_alpaca.yml +1 -1
- configs/llama_7B_4bit.yml +1 -1
- configs/llama_7B_alpaca.yml +1 -1
- configs/llama_7B_jeopardy.yml +1 -1
- configs/pythia_1_2B_alpaca.yml +1 -1
- configs/quickstart.yml +1 -1
- configs/sample.yml +1 -1
- configs/stability_3b.yml +1 -1
- configs/vicuna_13B_4bit_reflect.yml +1 -1
- examples/gptq-lora-7b/config.yml +1 -1
- examples/mpt-7b/config.yml +1 -1
- examples/redpajama/config-3b.yml +1 -1
configs/cerebras_1_3B_alpaca.yml
CHANGED
@@ -24,7 +24,7 @@ lora_fan_in_fan_out: false
|
|
24 |
wandb_project: pythia-1.4b-lora
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
-
wandb_log_model:
|
28 |
output_dir: ./lora-alpaca
|
29 |
batch_size: 32
|
30 |
micro_batch_size: 4
|
|
|
24 |
wandb_project: pythia-1.4b-lora
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
+
wandb_log_model:
|
28 |
output_dir: ./lora-alpaca
|
29 |
batch_size: 32
|
30 |
micro_batch_size: 4
|
configs/galactica_1_3B.yml
CHANGED
@@ -21,7 +21,7 @@ lora_fan_in_fan_out: false
|
|
21 |
wandb_project:
|
22 |
wandb_watch:
|
23 |
wandb_run_id:
|
24 |
-
wandb_log_model:
|
25 |
output_dir: ./lora-llama-alpaca
|
26 |
batch_size: 32
|
27 |
micro_batch_size: 16
|
|
|
21 |
wandb_project:
|
22 |
wandb_watch:
|
23 |
wandb_run_id:
|
24 |
+
wandb_log_model:
|
25 |
output_dir: ./lora-llama-alpaca
|
26 |
batch_size: 32
|
27 |
micro_batch_size: 16
|
configs/gpt_neox_20b.yml
CHANGED
@@ -23,7 +23,7 @@ lora_fan_in_fan_out: true # pythia/GPTNeoX lora specific
|
|
23 |
wandb_project: gpt4all-neox-20b
|
24 |
wandb_watch:
|
25 |
wandb_run_id:
|
26 |
-
wandb_log_model:
|
27 |
output_dir: ./gpt4all-neox-20b
|
28 |
batch_size: 48
|
29 |
micro_batch_size: 4
|
|
|
23 |
wandb_project: gpt4all-neox-20b
|
24 |
wandb_watch:
|
25 |
wandb_run_id:
|
26 |
+
wandb_log_model:
|
27 |
output_dir: ./gpt4all-neox-20b
|
28 |
batch_size: 48
|
29 |
micro_batch_size: 4
|
configs/llama_13B_alpaca.yml
CHANGED
@@ -21,7 +21,7 @@ lora_fan_in_fan_out: false
|
|
21 |
wandb_project:
|
22 |
wandb_watch:
|
23 |
wandb_run_id:
|
24 |
-
wandb_log_model:
|
25 |
output_dir: ./llama-13b-sharegpt
|
26 |
batch_size: 64
|
27 |
micro_batch_size: 2
|
|
|
21 |
wandb_project:
|
22 |
wandb_watch:
|
23 |
wandb_run_id:
|
24 |
+
wandb_log_model:
|
25 |
output_dir: ./llama-13b-sharegpt
|
26 |
batch_size: 64
|
27 |
micro_batch_size: 2
|
configs/llama_65B_alpaca.yml
CHANGED
@@ -27,7 +27,7 @@ lora_fan_in_fan_out: false
|
|
27 |
wandb_project: llama-65b-lora
|
28 |
wandb_watch:
|
29 |
wandb_run_id:
|
30 |
-
wandb_log_model:
|
31 |
output_dir: ./lora-llama-alpaca
|
32 |
batch_size: 128
|
33 |
micro_batch_size: 16
|
|
|
27 |
wandb_project: llama-65b-lora
|
28 |
wandb_watch:
|
29 |
wandb_run_id:
|
30 |
+
wandb_log_model:
|
31 |
output_dir: ./lora-llama-alpaca
|
32 |
batch_size: 128
|
33 |
micro_batch_size: 16
|
configs/llama_7B_4bit.yml
CHANGED
@@ -24,7 +24,7 @@ lora_fan_in_fan_out: false
|
|
24 |
wandb_project:
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
-
wandb_log_model:
|
28 |
output_dir: ./lora-test
|
29 |
batch_size: 8
|
30 |
micro_batch_size: 2
|
|
|
24 |
wandb_project:
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
+
wandb_log_model:
|
28 |
output_dir: ./lora-test
|
29 |
batch_size: 8
|
30 |
micro_batch_size: 2
|
configs/llama_7B_alpaca.yml
CHANGED
@@ -26,7 +26,7 @@ lora_fan_in_fan_out: false
|
|
26 |
wandb_project: llama-7b-lora
|
27 |
wandb_watch:
|
28 |
wandb_run_id:
|
29 |
-
wandb_log_model:
|
30 |
output_dir: ./lora-llama-alpaca
|
31 |
batch_size: 128
|
32 |
micro_batch_size: 16
|
|
|
26 |
wandb_project: llama-7b-lora
|
27 |
wandb_watch:
|
28 |
wandb_run_id:
|
29 |
+
wandb_log_model:
|
30 |
output_dir: ./lora-llama-alpaca
|
31 |
batch_size: 128
|
32 |
micro_batch_size: 16
|
configs/llama_7B_jeopardy.yml
CHANGED
@@ -22,7 +22,7 @@ lora_fan_in_fan_out: false
|
|
22 |
wandb_project: jeopardy-bot-7b
|
23 |
wandb_watch:
|
24 |
wandb_run_id:
|
25 |
-
wandb_log_model:
|
26 |
output_dir: ./jeopardy-bot-7b
|
27 |
batch_size: 4
|
28 |
micro_batch_size: 1
|
|
|
22 |
wandb_project: jeopardy-bot-7b
|
23 |
wandb_watch:
|
24 |
wandb_run_id:
|
25 |
+
wandb_log_model:
|
26 |
output_dir: ./jeopardy-bot-7b
|
27 |
batch_size: 4
|
28 |
micro_batch_size: 1
|
configs/pythia_1_2B_alpaca.yml
CHANGED
@@ -26,7 +26,7 @@ lora_fan_in_fan_out: true # pythia/GPTNeoX lora specific
|
|
26 |
wandb_project: pythia-1.4b-lora
|
27 |
wandb_watch:
|
28 |
wandb_run_id:
|
29 |
-
wandb_log_model:
|
30 |
output_dir: ./lora-alpaca
|
31 |
batch_size: 48
|
32 |
micro_batch_size: 4
|
|
|
26 |
wandb_project: pythia-1.4b-lora
|
27 |
wandb_watch:
|
28 |
wandb_run_id:
|
29 |
+
wandb_log_model:
|
30 |
output_dir: ./lora-alpaca
|
31 |
batch_size: 48
|
32 |
micro_batch_size: 4
|
configs/quickstart.yml
CHANGED
@@ -24,7 +24,7 @@ lora_fan_in_fan_out: false
|
|
24 |
wandb_project:
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
-
wandb_log_model:
|
28 |
output_dir: ./lora-test
|
29 |
batch_size: 4
|
30 |
micro_batch_size: 1
|
|
|
24 |
wandb_project:
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
+
wandb_log_model:
|
28 |
output_dir: ./lora-test
|
29 |
batch_size: 4
|
30 |
micro_batch_size: 1
|
configs/sample.yml
CHANGED
@@ -49,7 +49,7 @@ lora_fan_in_fan_out: false
|
|
49 |
wandb_project:
|
50 |
wandb_watch:
|
51 |
wandb_run_id:
|
52 |
-
wandb_log_model:
|
53 |
# where to save the finsihed model to
|
54 |
output_dir: ./completed-model
|
55 |
# training hyperparameters
|
|
|
49 |
wandb_project:
|
50 |
wandb_watch:
|
51 |
wandb_run_id:
|
52 |
+
wandb_log_model:
|
53 |
# where to save the finsihed model to
|
54 |
output_dir: ./completed-model
|
55 |
# training hyperparameters
|
configs/stability_3b.yml
CHANGED
@@ -20,7 +20,7 @@ lora_fan_in_fan_out: false
|
|
20 |
wandb_project: stable-alpaca-3b
|
21 |
wandb_watch:
|
22 |
wandb_run_id:
|
23 |
-
wandb_log_model:
|
24 |
output_dir: ./stable-alpaca-3b
|
25 |
batch_size: 2
|
26 |
micro_batch_size: 1
|
|
|
20 |
wandb_project: stable-alpaca-3b
|
21 |
wandb_watch:
|
22 |
wandb_run_id:
|
23 |
+
wandb_log_model:
|
24 |
output_dir: ./stable-alpaca-3b
|
25 |
batch_size: 2
|
26 |
micro_batch_size: 1
|
configs/vicuna_13B_4bit_reflect.yml
CHANGED
@@ -28,7 +28,7 @@ lora_fan_in_fan_out: false
|
|
28 |
wandb_project:
|
29 |
wandb_watch:
|
30 |
wandb_run_id:
|
31 |
-
wandb_log_model:
|
32 |
output_dir: ./lora-reflect
|
33 |
batch_size: 8
|
34 |
micro_batch_size: 2
|
|
|
28 |
wandb_project:
|
29 |
wandb_watch:
|
30 |
wandb_run_id:
|
31 |
+
wandb_log_model:
|
32 |
output_dir: ./lora-reflect
|
33 |
batch_size: 8
|
34 |
micro_batch_size: 2
|
examples/gptq-lora-7b/config.yml
CHANGED
@@ -24,7 +24,7 @@ lora_fan_in_fan_out: false
|
|
24 |
wandb_project: llama-7b-lora-int4
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
-
wandb_log_model:
|
28 |
output_dir: ./llama-7b-lora-int4
|
29 |
batch_size: 1
|
30 |
micro_batch_size: 1
|
|
|
24 |
wandb_project: llama-7b-lora-int4
|
25 |
wandb_watch:
|
26 |
wandb_run_id:
|
27 |
+
wandb_log_model:
|
28 |
output_dir: ./llama-7b-lora-int4
|
29 |
batch_size: 1
|
30 |
micro_batch_size: 1
|
examples/mpt-7b/config.yml
CHANGED
@@ -22,7 +22,7 @@ lora_fan_in_fan_out: false
|
|
22 |
wandb_project: mpt-alpaca-7b
|
23 |
wandb_watch:
|
24 |
wandb_run_id:
|
25 |
-
wandb_log_model:
|
26 |
output_dir: ./mpt-alpaca-7b
|
27 |
batch_size: 1
|
28 |
micro_batch_size: 1
|
|
|
22 |
wandb_project: mpt-alpaca-7b
|
23 |
wandb_watch:
|
24 |
wandb_run_id:
|
25 |
+
wandb_log_model:
|
26 |
output_dir: ./mpt-alpaca-7b
|
27 |
batch_size: 1
|
28 |
micro_batch_size: 1
|
examples/redpajama/config-3b.yml
CHANGED
@@ -23,7 +23,7 @@ lora_fan_in_fan_out: false
|
|
23 |
wandb_project: redpajama-alpaca-3b
|
24 |
wandb_watch:
|
25 |
wandb_run_id:
|
26 |
-
wandb_log_model:
|
27 |
output_dir: ./redpajama-alpaca-3b
|
28 |
batch_size: 4
|
29 |
micro_batch_size: 1
|
|
|
23 |
wandb_project: redpajama-alpaca-3b
|
24 |
wandb_watch:
|
25 |
wandb_run_id:
|
26 |
+
wandb_log_model:
|
27 |
output_dir: ./redpajama-alpaca-3b
|
28 |
batch_size: 4
|
29 |
micro_batch_size: 1
|