Upload folder using huggingface_hub
Browse files- checkpoint-1500/config.json +34 -0
- checkpoint-1500/generation_config.json +6 -0
- checkpoint-1500/model-00001-of-00002.safetensors +3 -0
- checkpoint-1500/model-00002-of-00002.safetensors +3 -0
- checkpoint-1500/model.safetensors.index.json +0 -0
- checkpoint-3000/config.json +34 -0
- checkpoint-3000/generation_config.json +6 -0
- checkpoint-3000/model-00001-of-00002.safetensors +3 -0
- checkpoint-3000/model-00002-of-00002.safetensors +3 -0
- checkpoint-3000/model.safetensors.index.json +0 -0
- checkpoint-4500/config.json +34 -0
- checkpoint-4500/generation_config.json +6 -0
- checkpoint-4500/model-00001-of-00002.safetensors +3 -0
- checkpoint-4500/model-00002-of-00002.safetensors +3 -0
- checkpoint-4500/model.safetensors.index.json +0 -0
- checkpoint-6000/config.json +34 -0
- checkpoint-6000/generation_config.json +6 -0
- checkpoint-6000/model-00001-of-00002.safetensors +3 -0
- checkpoint-6000/model-00002-of-00002.safetensors +3 -0
- checkpoint-6000/model.safetensors.index.json +0 -0
- checkpoint-7500/config.json +34 -0
- checkpoint-7500/generation_config.json +6 -0
- checkpoint-7500/model-00001-of-00002.safetensors +3 -0
- checkpoint-7500/model-00002-of-00002.safetensors +3 -0
- checkpoint-7500/model.safetensors.index.json +0 -0
checkpoint-1500/config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/phi-2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Flamingo"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "microsoft/phi-2--configuration_phi.PhiConfig",
|
| 9 |
+
"AutoModelForCausalLM": "microsoft/phi-2--modeling_phi.PhiForCausalLM"
|
| 10 |
+
},
|
| 11 |
+
"bos_token_id": 50256,
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"eos_token_id": 50256,
|
| 14 |
+
"hidden_act": "gelu_new",
|
| 15 |
+
"hidden_size": 2560,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 10240,
|
| 18 |
+
"layer_norm_eps": 1e-05,
|
| 19 |
+
"max_position_embeddings": 2048,
|
| 20 |
+
"model_type": "phi",
|
| 21 |
+
"num_attention_heads": 32,
|
| 22 |
+
"num_hidden_layers": 32,
|
| 23 |
+
"num_key_value_heads": 32,
|
| 24 |
+
"partial_rotary_factor": 0.4,
|
| 25 |
+
"qk_layernorm": false,
|
| 26 |
+
"resid_pdrop": 0.1,
|
| 27 |
+
"rope_scaling": null,
|
| 28 |
+
"rope_theta": 10000.0,
|
| 29 |
+
"tie_word_embeddings": false,
|
| 30 |
+
"torch_dtype": "bfloat16",
|
| 31 |
+
"transformers_version": "4.38.2",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 50298
|
| 34 |
+
}
|
checkpoint-1500/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.38.2"
|
| 6 |
+
}
|
checkpoint-1500/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fedecc731da839eedc9b5a33711dca9acdda96de6ec2bcd8751f18e5a6ab6b91
|
| 3 |
+
size 4949579540
|
checkpoint-1500/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e698f654489996460b0f0d7bc3e234a757d0c7bab7238d203a0116866ea819d7
|
| 3 |
+
size 4341749912
|
checkpoint-1500/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-3000/config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/phi-2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Flamingo"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "microsoft/phi-2--configuration_phi.PhiConfig",
|
| 9 |
+
"AutoModelForCausalLM": "microsoft/phi-2--modeling_phi.PhiForCausalLM"
|
| 10 |
+
},
|
| 11 |
+
"bos_token_id": 50256,
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"eos_token_id": 50256,
|
| 14 |
+
"hidden_act": "gelu_new",
|
| 15 |
+
"hidden_size": 2560,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 10240,
|
| 18 |
+
"layer_norm_eps": 1e-05,
|
| 19 |
+
"max_position_embeddings": 2048,
|
| 20 |
+
"model_type": "phi",
|
| 21 |
+
"num_attention_heads": 32,
|
| 22 |
+
"num_hidden_layers": 32,
|
| 23 |
+
"num_key_value_heads": 32,
|
| 24 |
+
"partial_rotary_factor": 0.4,
|
| 25 |
+
"qk_layernorm": false,
|
| 26 |
+
"resid_pdrop": 0.1,
|
| 27 |
+
"rope_scaling": null,
|
| 28 |
+
"rope_theta": 10000.0,
|
| 29 |
+
"tie_word_embeddings": false,
|
| 30 |
+
"torch_dtype": "bfloat16",
|
| 31 |
+
"transformers_version": "4.38.2",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 50298
|
| 34 |
+
}
|
checkpoint-3000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.38.2"
|
| 6 |
+
}
|
checkpoint-3000/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08472b71f51a6205dc91199731881cecdd19138659e75fee68e335d99877d64d
|
| 3 |
+
size 4949579540
|
checkpoint-3000/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c51b8123dec2cc562636ed64fbcdbf67792ed73e400e95b6c81ade7cee83a19c
|
| 3 |
+
size 4341749912
|
checkpoint-3000/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4500/config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/phi-2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Flamingo"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "microsoft/phi-2--configuration_phi.PhiConfig",
|
| 9 |
+
"AutoModelForCausalLM": "microsoft/phi-2--modeling_phi.PhiForCausalLM"
|
| 10 |
+
},
|
| 11 |
+
"bos_token_id": 50256,
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"eos_token_id": 50256,
|
| 14 |
+
"hidden_act": "gelu_new",
|
| 15 |
+
"hidden_size": 2560,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 10240,
|
| 18 |
+
"layer_norm_eps": 1e-05,
|
| 19 |
+
"max_position_embeddings": 2048,
|
| 20 |
+
"model_type": "phi",
|
| 21 |
+
"num_attention_heads": 32,
|
| 22 |
+
"num_hidden_layers": 32,
|
| 23 |
+
"num_key_value_heads": 32,
|
| 24 |
+
"partial_rotary_factor": 0.4,
|
| 25 |
+
"qk_layernorm": false,
|
| 26 |
+
"resid_pdrop": 0.1,
|
| 27 |
+
"rope_scaling": null,
|
| 28 |
+
"rope_theta": 10000.0,
|
| 29 |
+
"tie_word_embeddings": false,
|
| 30 |
+
"torch_dtype": "bfloat16",
|
| 31 |
+
"transformers_version": "4.38.2",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 50298
|
| 34 |
+
}
|
checkpoint-4500/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.38.2"
|
| 6 |
+
}
|
checkpoint-4500/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4034a5e4273a8324c9b1c3067e7dc6ae00e7291888d338671a3ca13e48ea3f1f
|
| 3 |
+
size 4949579540
|
checkpoint-4500/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a11dbb3706c793f11bd3108073090a2e6c4a620184ccd3b0af528955e8a51a1d
|
| 3 |
+
size 4341749912
|
checkpoint-4500/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-6000/config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/phi-2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Flamingo"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "microsoft/phi-2--configuration_phi.PhiConfig",
|
| 9 |
+
"AutoModelForCausalLM": "microsoft/phi-2--modeling_phi.PhiForCausalLM"
|
| 10 |
+
},
|
| 11 |
+
"bos_token_id": 50256,
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"eos_token_id": 50256,
|
| 14 |
+
"hidden_act": "gelu_new",
|
| 15 |
+
"hidden_size": 2560,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 10240,
|
| 18 |
+
"layer_norm_eps": 1e-05,
|
| 19 |
+
"max_position_embeddings": 2048,
|
| 20 |
+
"model_type": "phi",
|
| 21 |
+
"num_attention_heads": 32,
|
| 22 |
+
"num_hidden_layers": 32,
|
| 23 |
+
"num_key_value_heads": 32,
|
| 24 |
+
"partial_rotary_factor": 0.4,
|
| 25 |
+
"qk_layernorm": false,
|
| 26 |
+
"resid_pdrop": 0.1,
|
| 27 |
+
"rope_scaling": null,
|
| 28 |
+
"rope_theta": 10000.0,
|
| 29 |
+
"tie_word_embeddings": false,
|
| 30 |
+
"torch_dtype": "bfloat16",
|
| 31 |
+
"transformers_version": "4.38.2",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 50298
|
| 34 |
+
}
|
checkpoint-6000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.38.2"
|
| 6 |
+
}
|
checkpoint-6000/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6cba53c7c3fc327104e587289db5ec9de73bbfead246d5fc8e55b6f6e9ce7fb1
|
| 3 |
+
size 4949579540
|
checkpoint-6000/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a34c29b03fdcdd6fb309d40c20eb195a07aca77525b24828681b19de3ac8ade3
|
| 3 |
+
size 4341749912
|
checkpoint-6000/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-7500/config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/phi-2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Flamingo"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"auto_map": {
|
| 8 |
+
"AutoConfig": "microsoft/phi-2--configuration_phi.PhiConfig",
|
| 9 |
+
"AutoModelForCausalLM": "microsoft/phi-2--modeling_phi.PhiForCausalLM"
|
| 10 |
+
},
|
| 11 |
+
"bos_token_id": 50256,
|
| 12 |
+
"embd_pdrop": 0.0,
|
| 13 |
+
"eos_token_id": 50256,
|
| 14 |
+
"hidden_act": "gelu_new",
|
| 15 |
+
"hidden_size": 2560,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 10240,
|
| 18 |
+
"layer_norm_eps": 1e-05,
|
| 19 |
+
"max_position_embeddings": 2048,
|
| 20 |
+
"model_type": "phi",
|
| 21 |
+
"num_attention_heads": 32,
|
| 22 |
+
"num_hidden_layers": 32,
|
| 23 |
+
"num_key_value_heads": 32,
|
| 24 |
+
"partial_rotary_factor": 0.4,
|
| 25 |
+
"qk_layernorm": false,
|
| 26 |
+
"resid_pdrop": 0.1,
|
| 27 |
+
"rope_scaling": null,
|
| 28 |
+
"rope_theta": 10000.0,
|
| 29 |
+
"tie_word_embeddings": false,
|
| 30 |
+
"torch_dtype": "bfloat16",
|
| 31 |
+
"transformers_version": "4.38.2",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 50298
|
| 34 |
+
}
|
checkpoint-7500/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.38.2"
|
| 6 |
+
}
|
checkpoint-7500/model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53d720eed872c5ff6e73f9a5e8c7e20ef8626feff6cb9b0bf8fc15056dc1c75d
|
| 3 |
+
size 4949579540
|
checkpoint-7500/model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74d004bdea14bc909a9796898e05b349de4d04be3842dffc4777c829f82367aa
|
| 3 |
+
size 4341749912
|
checkpoint-7500/model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|