revert some changes now that caching is fixed

axolotl-ai-cloud · Jan 9, 2024 · a8d3c31 · a8d3c31
1 parent f01ca4f
commit a8d3c31
Show file tree

Hide file tree

Showing 12 changed files with 2 additions and 21 deletions.
diff --git a/docker/Dockerfile-tests b/docker/Dockerfile-tests
@@ -24,9 +24,9 @@ RUN git fetch origin +$GITHUB_REF && \
 
 # If AXOLOTL_EXTRAS is set, append it in brackets
 RUN if [ "$AXOLOTL_EXTRAS" != "" ] ; then \
-        pip install -e .[deepspeed,flash-attn,$AXOLOTL_EXTRAS]; \
+        pip install -e .[deepspeed,flash-attn,mamba-ssm,$AXOLOTL_EXTRAS]; \
     else \
-        pip install -e .[deepspeed,flash-attn]; \
+        pip install -e .[deepspeed,flash-attn,mamba-ssm]; \
     fi
 
 # So we can test the Docker image

diff --git a/tests/e2e/test_fused_llama.py b/tests/e2e/test_fused_llama.py
@@ -31,7 +31,6 @@ def test_fft_packing(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "JackFram/llama-68m",
                 "flash_attention": True,
                 "flash_attn_fuse_qkv": True,

diff --git a/tests/e2e/test_lora_llama.py b/tests/e2e/test_lora_llama.py
@@ -31,7 +31,6 @@ def test_lora(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "JackFram/llama-68m",
                 "tokenizer_type": "LlamaTokenizer",
                 "sequence_len": 1024,
@@ -74,7 +73,6 @@ def test_lora_packing(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "JackFram/llama-68m",
                 "tokenizer_type": "LlamaTokenizer",
                 "sequence_len": 1024,
@@ -124,7 +122,6 @@ def test_lora_gptq(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "TheBlokeAI/jackfram_llama-68m-GPTQ",
                 "model_type": "AutoModelForCausalLM",
                 "tokenizer_type": "LlamaTokenizer",

diff --git a/tests/e2e/test_mamba.py b/tests/e2e/test_mamba.py
@@ -29,7 +29,6 @@ def test_fft(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "state-spaces/mamba-130m",
                 "model_type": "MambaLMHeadModel",
                 "tokenizer_type": "AutoTokenizer",

diff --git a/tests/e2e/test_mistral.py b/tests/e2e/test_mistral.py
@@ -31,7 +31,6 @@ def test_lora(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "openaccess-ai-collective/tiny-mistral",
                 "flash_attention": True,
                 "sequence_len": 1024,
@@ -77,7 +76,6 @@ def test_ft(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "openaccess-ai-collective/tiny-mistral",
                 "flash_attention": True,
                 "sequence_len": 1024,

diff --git a/tests/e2e/test_mistral_samplepack.py b/tests/e2e/test_mistral_samplepack.py
@@ -31,7 +31,6 @@ def test_lora_packing(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "openaccess-ai-collective/tiny-mistral",
                 "flash_attention": True,
                 "sample_packing": True,
@@ -78,7 +77,6 @@ def test_ft_packing(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "openaccess-ai-collective/tiny-mistral",
                 "flash_attention": True,
                 "sample_packing": True,

diff --git a/tests/e2e/test_mixtral.py b/tests/e2e/test_mixtral.py
@@ -31,7 +31,6 @@ def test_qlora(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "hf-internal-testing/Mixtral-tiny",
                 "tokenizer_config": "mistralai/Mixtral-8x7B-v0.1",
                 "flash_attention": True,
@@ -74,7 +73,6 @@ def test_ft(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "hf-internal-testing/Mixtral-tiny",
                 "tokenizer_config": "mistralai/Mixtral-8x7B-v0.1",
                 "flash_attention": True,

diff --git a/tests/e2e/test_mixtral_samplepack.py b/tests/e2e/test_mixtral_samplepack.py
@@ -31,7 +31,6 @@ def test_qlora(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "hf-internal-testing/Mixtral-tiny",
                 "tokenizer_config": "mistralai/Mixtral-8x7B-v0.1",
                 "flash_attention": True,
@@ -79,7 +78,6 @@ def test_ft(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "hf-internal-testing/Mixtral-tiny",
                 "tokenizer_config": "mistralai/Mixtral-8x7B-v0.1",
                 "flash_attention": True,

diff --git a/tests/e2e/test_model_patches.py b/tests/e2e/test_model_patches.py
@@ -21,7 +21,6 @@ class TestModelPatches(unittest.TestCase):
     def test_mixtral_multipack(self, temp_dir):
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "hf-internal-testing/Mixtral-tiny",
                 "tokenizer_config": "mistralai/Mixtral-8x7B-v0.1",
                 "flash_attention": True,
@@ -65,7 +64,6 @@ def test_mixtral_multipack(self, temp_dir):
     def test_mistral_multipack(self, temp_dir):
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "openaccess-ai-collective/tiny-mistral",
                 "flash_attention": True,
                 "sample_packing": True,

diff --git a/tests/e2e/test_phi.py b/tests/e2e/test_phi.py
@@ -33,7 +33,6 @@ def test_phi2_ft(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "microsoft/phi-2",
                 "trust_remote_code": True,
                 "model_type": "AutoModelForCausalLM",
@@ -85,7 +84,6 @@ def test_ft_packed(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "microsoft/phi-2",
                 "trust_remote_code": True,
                 "model_type": "PhiForCausalLM",

diff --git a/tests/e2e/test_resume.py b/tests/e2e/test_resume.py
@@ -33,7 +33,6 @@ def test_resume_qlora(self, temp_dir):
         # pylint: disable=duplicate-code
         cfg = DictDefault(
             {
-                "wandb_disabled": True,
                 "base_model": "JackFram/llama-68m",
                 "tokenizer_type": "LlamaTokenizer",
                 "sequence_len": 1024,

diff --git a/tests/test_validation.py b/tests/test_validation.py
@@ -721,7 +721,6 @@ def test_add_tokens_adapter(self):
         validate_config(cfg)
 
 
-@pytest.mark.skip(reason="skip wandb validation for docker e2e")
 class ValidationWandbTest(ValidationTest):
     """
     Validation test for wandb